United States      Office of Air Quality
        Environmental Protection Planning and Standards
        Agency         Research Triangle Park NC 27711
EPA-453/R-94-O03a
March 1994
        Air
EPA     Hazardous Air Pollutant          Final
        Emissions from Process         EIS
        Units in the
        Synthetic Organic Chemical
        Manufacturing Industry-
        Background Information
        for Final Standards

        Volume 2A:  Comments on
                     Process Vents,
                     Storage Vessels,
                     Transfer Operations,
                     and Equipment Leaks

-------
                      EPA-453/R-94-003a
 Hazardous Air Pollutant Emissions
     from Process Units in the
    Synthetic Organic Chemical
     Manufacturing Industry--
      Background Information
        for Final Standards
      Volume 2A:  Comments on
    Process Vents, Storage Vessels,
       Transfer Operations, and
           Equipment Leaks

     Emission Standards Division
    U.S. Environmental Protection Agency
        Office of Air and Radiation
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
March 1 994       '   6^o4. i2ttl
                                         Floae

-------
                          DISCLAIMER
This Report has been reviewed by the Emission Standards
Division of the Office of Air Quality Planning and Standards,
EPA, and approved for publication.  Mention of trade names or
commercial products is not intended to constitute endorsement
or recommendation for use.  Copies of this report are
available through the Library Services Office (MD-35),  U.S.
Environmental Protection Agency, Research Triangle Park, NC
27711, or from the National Technical Information Service,
5285 Port Royal Road, Springfield, VA  22161.

-------
                ENVIRONMENTAL PROTECTION AGENCY
        Background Information and Final  Environmental
         Impact Statement .for Hazardous Air Pollutant
        Emissions  from Process Units in the Synthetic
            Organic Chemical Manufacturing Industry
    Volume 2A:   Comments  on Process Vents,  Storage Vessels,
              Transfer Racks,  and Equipment Leaks
                         Prepared by:
Bruce ^Pordan                                         (Date)
Director, Emission Standards Division
U.S. Environmental Protection Agency
Research Triangle Park, NC  27711

1.      The  standards  regulate  emissions  of  organic  hazardous
        air  pollutants (HAP's)  emitted  from  chemical
        manufacturing  process units  in  the Synthetic Organic
        Chemical  Manufacturing  Industry (SOCMI)  and  from  other
        processes subject  to the  negotiated  regulation  for
        equipment leaks.   Only  those  chemical  manufacturing
        process units  that are  part  of  major sources under
        Section 112(d)  of  the Clean  Air Act  (Act)  will  be
        regulated.  The standards will  reduce  emissions of  112
        of the organic chemicals  identified  in the Act  list of
        189  HAP's.

2.      Copies of this document have  been sent to  the following
        Federal Departments:  Labor,  Health  and Human Services,
        Defense,  Transportation,  Agriculture,  Commerce,
        Interior,  and  Energy; the National Science Foundation;
        and  the Council on Environmental  Quality.  Copies have
        also been sent to  members of  the  State and Territorial
        Air  Pollution  Program Administrators;  the  Association
        of Local  Air Pollution  Control  Officials;  EPA Regional
        Administrators; and other interested parties.

3.      For  additional information contact:

        Dr.  Janet Meyer
        Standards Development Branch  (MD-13)
        U.S. Environmental Protection Agency
        Research  Triangle  Park, NC   27711
        Telephone:   (919)  541-5254

-------
4.       Paper copies of this document may be obtained  from:

        National Technical  Information Service  (NTIS)
        5285 Port Royal Road
        Springfield, VA   22161
        Telephone:  (703)  487-4650

5.       Electronic  copies of this document may  be  obtained  from
        the EPA Technology  Transfer Network  (TTN).   The  TTN is
        an electronic bulletin board system which  is free,
        except for  the normal long distance charges.   To access
        the HON BID:

        •    Set software to data bits:  8, N;  stop  bits:   1
        •    Use access number  (919) 541-5742 for  1200,  2400,
            or 9600 bps  modems  [access  problems  should  be
            directed to  the system operator at (919)
            541-5384].
        •    Specify TTN  Bulletin Board:  Clean Air  Act
            Amendments
        •    Select menu  item:   Recently Signed Rules

-------
                           OVERVIEW

     Emission standards under section 112(d) of the Clean Air
Act  (Act) apply to new and existing sources in each listed
category of hazardous air pollutant (HAP) emission sources.
This background information document (BID)  provides summaries
and responses for public comments received regarding the
Hazardous Organic National Emission Standard for Hazardous Air
Pollutants (NESHAP), commonly referred to as the HON.  The HON
will primarily affect the Synthetic Organic Chemical
Manufacturing Industry (SOCMI).  However, the provisions for
equipment leaks also apply to certain polymer and resin
production processes, certain pesticide production processes,
and certain miscellaneous processes that are subject to the
negotiated regulation for equipment leaks.
     This BID comprises six volumes as follows,:
     •    Volume 2A:  Comments on Process Vents, Storage
          Vessels, Transfer Operations, and Equipment Leaks
          (EPA-453/R-94-003a);
     •    Volume 2B:  Comments on Wastewater
          (EPA-453/R-94-003b);
     •    Volume 2C:  Comments on Emissions Averaging
          (EPA-453/R-94-003C);
     •    Volume 2D: , Comments on Applicability, National
          Impacts, and Overlap with Other Rules
          (EPA-453/R-94-003d);
     •    Volume 2E:  Comments on Recordkeeping, Reporting,
          Compliance, and Test Methods (EPA-453/R-94-003e);
          and .
     •    Volume 2F:  Commenter Identification List
          (EPA-453/R-94-003f).

     Volume 2A is organized by emission point and contains
discussions of specific technical issues related to process

-------
vents, storage vessels,  transfer operations,  and equipment
leaks.  Volume 2A discusses specific technical issues such as
control technology, cost analysis,  emission estimates,
Group I/Group 2 determination,  compliance options and
demonstrations, and monitoring.
     Volume 2B addresses issues related to controlling
emissions from wastewater.   Specific technical issues include
control technology, cost analysis,  emission estimates,
Group I/Group 2 determination,  compliance options and
demonstrations, and monitoring.
     Volume 2C contains the EPA's decisions regarding
emissions averaging.  Specific issues include the scope of
emissions averaging in the HON, specific provisions related to
credits and banking, and enforcement of an emissions averaging
system for the HON.
     Volume 2D discusses applicability of the HON in terms of
selection of source category, selection of source, and
selection of pollutants.  Volume 2D also addresses the process
for determination of the MACT floor and selection of the
specific applicability thresholds for process vents, storage
vessels, transfer racks, wastewater operations, and equipment
leaks.
     Volume 2E discusses the provisions for compliance,
recordkeeping and reporting.  Volume 2E also discusses issues
related to the use of EPA test methods.
     Volume 2F of each volume contains a list of commenters,
their affiliations, and the EPA docket and item number
assigned to each comment.

-------
                       TABLE OF CONTENTS

Section                   .                                 Page
ACRONYM AND ABBREVIATION LIST	     X
LIST OF FREQUENTLY USED TERMS	   xiv
1.0  INTRODUCTION	   1-1
2.0  PROCESS VENTS	   2-1
     2.1  EMISSION CONTROL TECHNOLOGY 	   2-1
     2.2  IMPACTS ANALYSIS  	  2-12
          2.2.1  Cost Analysis	2-12
          2.2.2  Emission Estimates   	  2-15
     2.3  APPLICABILITY AND GROUP I/GROUP 2
          DETERMINATION 	  2-15
          2.3.1  Applicability	2-15
          2.3.2  Group I/Group 2 Determination  	  2-18
     2.4  COMPLIANCE DEMONSTRATIONS 	  2-27
          2.4.1  Performance Testing  	  2-27
          2.4.2  Monitoring   	2-34
     2.5  WORDING OF THE PROVISIONS	2-45
3.0  STORAGE VESSELS	   3-1
     3.1  EMISSION CONTROL TECHNOLOGY 	   3-1
     3.2  IMPACTS ANALYSIS  	  3-10
          3.2.1  Cost Impacts   	3-10
          3.2.2  Emission Estimates   	  3-16
     3.3  APPLICABILITY AND GROUP I/GROUP 2
          DETERMINATION 	  3-17
          3.3.1  Applicability	3-17
          3.3.2  Group I/Group 2 Determination  	  3-21
     3.4  COMPLIANCE	3-25
          3.4.1  General	3-25
          3.4.2  Routine Maintenance  	  3-28
          3.4.3  Compliance Schedule  	  3-33
          3.4.4  Inspections and Delay  of Repair  ....  3-37
     3.5  RECORDKEEPING AND REPORTING 	  3-45
     3.6  WORDING OF THE PROVISIONS	3-52
     3.7  MISCELLANEOUS	3-55
                              Vll

-------
                 TABLE OF CONTENTS, CONTINUED

Section                                                   Page
4.0  TRANSFER OPERATIONS  	   4-1
     4.1  APPLICABILITY AND GROUP I/GROUP 2
          DETERMINATION	   4-1
          4.1.1  Applicability	   4-5
          4.1.2  Group I/Group 2 Determination  	   4-8
     4.2  COMPLIANCE	4-11
          4.2.1  Performance Testing  	 4-11
          4.2.2  Monitoring   	4-19
          4.2.3  Inspections	4-21
          4.2.4  Compliance Schedule  	 4-26
     4.3  RECORDKEEPING AND REPORTING 	 4-26
     4.4  WORDING OF THE PROVISIONS	4-29
     4.5  MISCELLANEOUS	4-30
5.0  EQUIPMENT LEAKS	   5-1
     5.1  STANDARDS	   5-1
          5.1.1  §63.162:  General  	   5-1
          5.1.2  §63.163:  Pumps in Light  Liquid Service   5-7
          5.1.3  §63.164:  Compressors  	 5-10
          5.1.4  §63.165:  Pressure Relief Valves in
                 Gas/Vapor Service  	  .... 5-12
          5.1.5  §63.166:  Sampling Connection Systems   . 5-15
          5.1.6  §63.167:  Open-ended Valves or Lines    . 5-18
          5.1.7  §63.168:  Valves  in Gas/Vapor Service
                 and in Light Liquid Service	5-19
          5.1.8  §63.169:  Pumps, Valves,  Connectors, and
                 Agitators in Heavy Liquid Service;
                 Instrumentation Systems;  and Pressure
                 Relief Valves in  Liquid Service  .... 5-27
          5.1.9  §63.170:  Product Accumulator Vessels   . 5-28
          5.1.10 §63.171:  Delay of Repair 	 5-29
          5.1.11 §63.172:  Closed-vent Systems and
                 Control Devices  	 5-32
                             Vlll

-------
                 TABLE OF CONTENTS, CONTINUED

Section                                                    Page
          5.1.12 §63.173:  Agitators in Gas/Vapor Service
                 and in Light Liquid Service	5-34
          5.1.13 §63.174:  Connectors in Gas/Vapor
                 Service and in Light Liquid Service  .  .  5-35
          5.1.14 §63.175:  Quality Improvement Program
                 for Valves   	5-42
          5.1.15 §63.176:  Quality Improvement Program
                 for Pumps	5-45
          5.1.16 §63.177:  Alternative Means of Emission
                 Limitation   	5-45
          5.1.17 §63.178:  Alternative Means of Emission
                 Limitation for Batch Processes   ....  5-46
          5.1.18 §63.179:  Alternative Means of Emission
                 Limitation for Enclosed-Vented Process
                 Units	5-50
          5.1.19 Repair Procedures  	  5-50
     5.2  IMPACTS ANALYSIS  	  5-52
     5.3  APPLICABILITY	5-52
          5.3.1  Definition of SOCMI	5-52
          5.3.2  Definition of Non-SOCMI Processes  .  .  .  5-54
          5.3.3  Equipment Subject to Subpart H   ....  5-57
          5.3.4  Miscellaneous	5-58
     5.4  COMPLIANCE DEMONSTRATIONS 	  5-61
          5.4.1  Monitoring   	5-61
          5.4.2  Test Methods   	5-64
     5.5  RECORDKEEPING AND REPORTING 	  5-64
          5.5.1  General	5-64
          5.5.2  Batch Processes	5-67
     5.6  WORDING OF PROVISIONS	5-68
                              IX

-------
           ACRONYM AND ABBREVIATION LIST
ACRONYM
Act
ALAPCO

ASPEN

BACT
BAT
BD
BID
BIF
CEM
CFR
CMA
CMPU
CO
CTG
CWA
DMS
DOT
ORE
EB/S
EDC
EFR
EO
E.O.
EPA
Fe
Fm
FR
Fr
FTIR
HAP
TERM
Clean Air Act
Association of Local Air Pollution
Control Officers
advanced system for process
engineering
best available control technology
best available technology
butadiene
background information document
boilers and industrial furnaces
continuous emissions monitoring
Code of Federal Regulations
Chemical Manufacturers Association
chemical manufacturing process unit
carbon monoxide
control techniques guideline
Clean Water Act
dual mechanical seal
Department of Transportation
destruction and removal efficiency
ethylbenzene/styrene
ethylene dichloride
external floating roof
ethylene oxide
Executive Order
Environmental Protection Agency
fraction emitted
fraction measured
FEDERAL REGISTER
fraction removed
Fourier transform infrared
hazardous air pollutant

-------
      ACRONYM

ACRONYM

HON
IFR
LDAR
LAER
MACT
MIBK
MR
NCS
NESHAP

NOX
NPDES

NRDC
NSPS
NSR
OCCM

OCPSF

OMB
OSHA

P.L.
PAV
POM
POTW
PRA
PRV
PSD

QIP
AND ABBRIEVIATION LIST,  CONTINUED
      TERM
      hazardous organic national.emission
      standards for hazardous air
      pollutants
      internal floating roof
      leak detection and repair
      lowest achievable emission rate
      maximum achievable control technology
      methyl isobutyl ketone
      mass removal (actual)
      Notification of Compliance Status
      national emission standards for
      hazardous air pollutants
      nitrogen oxides
      National Pollutant Discharge
      Elimination System
      Natural Resources Defense Council
      new source performance standards
      new source review
      Office of Air Quality Planning and
      Standards Control Cost Manual
      organic chemicals, plastics, and
      synthetic fibers
      Office of Management and Budget
      Occupational Safety and Health
      Administration
      Public Law
      product accumulator vessel
      polycyclic organic matter
      publicly owned treatment works
      Paperwork Reduction Act
      pressure relief valve
      prevention of significant
      deterioration
      quality improvement program
                         xi

-------
      ACRONYM
ACRONYM
R & 0
RCRA

RCT
RIA
RMR
SARA

SIP
SMS
SOCMI

STAPPA

TAG
TACB
TCI
THC
TIC
TOC
TRE
TRI
TSDF

VHAP
VO

VOC
VOHAP
AND ABBREVIATION LIST, CONTINUED
      TERM
      research and development
      Resource Conservation and Recovery
      Act
      reference control technology
      Regulatory Impact Analysis
      required mass removal
      Superfund Amendment and
      Reauthorization Act
      State Implementation Plan
      single mechanical seal
      synthetic organic chemical
      manufacturing industry
      State and Territorial Air Pollution
      Program Administrators
      total annual cost
      Texas Air Control Board
      total capital investment
      total hydrocarbon
      total industry control
      total organic compound
      total resource effectiveness
      toxics release inventory
      treatment, storage, and disposal
      facility
      volatile hazardous air pollutant
      volatile organics measurable  by
      Method 25D
      volatile organic compound
      volatile organic hazardous  air
      pollutant
ABBREVIATION
bbl
BOB
      UNIT OF
      barrel
      barrels of oil  equivalent
                         xii

-------
      ACRONYM
ABBREVIATION
Btu
BtU/kW-hr

°C
oF
gal
gpm
hr
kg/hr
kPa
kW-hr/yr
£/hour» m2
£pm
gal
iP
Mg
mg
mg/dscm

MW
ppb
ppm
ppmv
ppmw
psia
scm/min
TJ
yr
AND ABBREVIATION LIST,  CONTINUED
      UNIT OF MEASURE
      British thermal unit
      British thermal unit per
      kilowatt-hour
      degrees Celsius
      degrees Fahrenheit
      gallon
      gallons per minute
      hour
      kilograms per hour
      kilopascals
      kilowatt-hour per year
      liters per hour per square meter
      liters per minute
      gallons
      cubic meters
      megagrams
      milligrams
      milligram per dry standard cubic
      meter
      megawatts
      parts per billion
      parts per million
      parts per million by volume
      parts per million by weight
      pounds per square inch absolute
      standard cubic meter per minute
      terajoules
      year
                        xiii

-------
                 LIST OF FREQUENTLY USED TERMS

Act means the Clean Air Act as amended in 1990.

Administrator means the Administrator of the U. S.
Environmental Protection Agency or his or her authorized
representative  (e.g., a State that has been delegated the
authority to implement the provisions of part 63).

Enhanced monitoring rule means the rule to be located in
sections 64.1 through 64.9 of part 64 of title 40 of the Code
of Federal Regulations.  This rule implements section 702(b)
of title VII of the 1990 Clean Air Act Amendments.  This rule
establishes the criteria and procedures that owners or
operators must satisfy in evaluating, selecting and
demonstrating enhanced monitoring, and includes appendices
containing enhanced monitoring performance and quality
assurance requirements.  The enhanced monitoring rule does not
apply to sources subject to 40 CFR part 63, and therefore does
not apply to sources subject to the HON.  The proposed rule
was published in the Federal Register on October 22, 1993
(58 FR 54648).

General Provisions means the general provisions located in
subpart A of part 63 of title 40 of the Code of Federal
Regulations.  These General Provisions codify national
emission standards for hazardous air pollutants (NESHAP) for
source categories covered under section 112 of the Act as
amended November 15, 1990.

Implementing aaencv means the Administrator of the U. S.
Environmental Protection Agency or a State, federal, or other
agency that has been delegated the authority to implement the
provisions of part 63.  Under section 112(1) of the Act,
States and localities may develop and submit to the
Administrator for approval a program for the implementation
and enforcement of emission standards.  A program submitted by
                              xiv

-------
            LIST OF FREQUENTLY USED TERMS,  CONTINUED
the State under section  112(1) of  the Act may provide for
partial or  complete delegation of  the Administrator's
authorities and responsibilities to implement and enforce
emission standards.

Operating permit program rule  means the rule located in
sections 70,1 through 70.11 of part 70 of chapter I of
title 40 of the Code of  Federal Regulations.  This  rule
implements  section 502(b) of title V of the 1990 Clean Air Act
Amendments.  Under this  rule,  States are required to develop,
and to submit to the EPA, programs for issuing operating
permits to  major stationary sources (including major sources
of hazardous air pollutants listed in section 112 of the Act),
sources covered by New Source  Performance Standards (NSPS),
sources covered by emissions standards for hazardous air
pollutants  pursuant to section 112 of the Act, and  affected
sources under the acid rain program.  The final rule was
published in the Federal Register  on July 21, 1992
(57 FR 32250).

Permitting  authority means:   (1) the State air pollution
control agency,  local agency,  other State agency, or other
agency authorized by the Administrator to carry out a permit
program under part 70; or (2)  the Administrator, in the case
of EPA-implemented permit programs under part 71.

Section 112 fa)   rule means the  rule to be located in subpart B
of part 63  of title 40 of the  Code of Federal Regulations.
This rule implements section 112(g) of the 1990 Clean Air Act
Amendments.  This rule will impose control technology
requirements on  "constructed,  reconstructed or modified" major
sources of hazardous air pollutants not already regulated by a
section 112(d)  or 112(j)  MACT  standard.
                              XV

-------
           LIST OF FREQUENTLY USED TERMS, CONTINUED
Section 112(1) rule means the rule located in subpart E of
part 63 of title 40 of the Code of Federal Regulations.  Under
this rule, a State or locality may submit a program to the
Administrator to request partial or complete delegation of the
Administrator's authorities and responsibilities to implement
and enforce section 112 emission standards.  The final rule
was published in the Federal Register on November 26, 1993
(58 FR 62262) .

Title III means title III of the 1990 Clean Air Act
Amendments.  Section 112 of the Act authorizes the EPA to
establish MACT standards.

Title V means title V of the 1990 Clean Air Act Amendments,
which authorizes the EPA to establish the operating permit
program.

Title VII means title VII of the 1990 Clean Air Act
Amendments. Section 702(b) of the Act authorizes the EPA to
establish compliance certification procedures.  The part 64
enhanced monitoring rule implements section 702(b).
                              xvi

-------
                       1.0  INTRODUCTION

     On December 31, 1992, the U.S. Environmental Protection
Agency (EPA) proposed the Hazardous Organic National Emission
Standard for Hazardous Air Pollutants (NESHAP) for process
units in the synthetic organic chemical manufacturing industry
(SOCMI) under section 112(d) of the Clean Air Act
(57 FR 62608).  Public comments were requested on the proposed
standard and comment letters were received from industry
representatives, governmental entities, environmental groups,
and private citizens.  Two public hearings were held, one in
Research Triangle Park (RTF), North Carolina, on February 25,
1993, and another in Baton Rouge, Louisiana, on
March 18, 1993.  Both hearings were open to the public and
5 persons in RTP and 45 persons in Baton Rouge presented oral
testimony on the proposed NESHAP.
     On August 11, 1993, the General Provisions for part 63
(58 FR 42760) were proposed.  In order to allow the public to
comment on how the General Provisions relate to the Hazardous
Organic NESHAP (HON), a supplemental notice (October 15, 1993;
58 FR 53478) was published.  Public comments were requested on
the overlap between the General Provisions and the HON and on
some specific emissions averaging issues.  Comment letters
regarding the supplemental notice were received from
80 commenters.
     The written comments that were submitted and verbal
comments made at the public hearings regarding the technical
and policy issues associated with process vents,  storage
vessels,  transfer operations, and equipment leaks in the
proposed rule and supplemental notice, along with responses to
these comments, are summarized in the following chapters.  The
                              1-1

-------
comments that were submitted regarding process vents are
summarized in chapter 2.0 and the comments regarding storage
vessels are summarized in chapter 3.0.  The comments that were
submitted regarding transfer operations and equipment leaks
are summarized in chapters 4.0 and 5.0, respectively.  The
summary of comments and responses serves as the basis for the
revisions made to the NESHAP between proposal and
promulgation.
     Within each chapter, the comments are organized into
sections such as:  emission control technology; impacts
analysis; applicability and Group I/Group 2 determination;
compliance demonstrations; recordkeeping and reporting;
wording of the provisions; and miscellaneous.  The emission
control technology section focuses on comments regarding the
applicability and performance of the reference control
technologies.  The impacts analysis section addresses comments
concerning cost analysis, emissions estimates, other
environmental impacts, and energy impacts.  The applicability
and Group 1/Group 2 determination section addresses comments
on the emission points covered by the NESHAP as well as which
emission points should be required to apply control.  The
compliance demonstrations section focuses on performance
testing, design evaluations, inspections, and monitoring.  The
reporting and recordkeeping section addresses comments
relating to the specific emission sources.  General
recordkeeping and reporting issues are presented in BID
Volume 2E.  The wording of the provisions section addresses
comments concerning clarification or consistency of the NESHAP
requirements and definitions, and the miscellaneous section
covers comments which did not fit in the other sections.
                              1-2

-------
                       2.0  PROCESS VENTS

2.1  EMISSION CONTROL TECHNOLOGY
     Comment:  One commenter (A-90-19:  IV-D-32) supports the
RCT sited for process vents, but requested that further
consideration be given to the requirements of vents with
existing control devices with a ORE between 95 and 98 percent.
Two conunenters (A-90-19:  IV-D-32; and IV-D-112) suggested
that facilities with catalytic incinerators achieving
95 percent ORE be allowed to continue to operate for a period
of time (e.g., 10 years) or until replacement is necessary.
Another commenter (A-90-19:  IV-D-97) advocated a 95 percent
control for existing facilities for a period of 10 years.  The
commenter (A-90-19:  IV-D-32) acknowledged the possibility of
using emissions averaging to make up the short fall between
95 percent and 98 percent control, but claimed that emissions
averaging may not be a viable option, especially for a small
facility,  and alternatives should therefore be made available.
     Response;  Existing process vent control devices that are
used to comply with the distillation or air oxidation NSPS or
State regulations in Texas, California, Illinois, and
Louisiana are required to have a removal efficiency of
98 percent.   The Ohio regulation also requires 98 percent
control of all air oxidation vents that are subject to the
regulation.   Therefore, the EPA has concluded that there are
few existing control devices for process vents that are
achieving a control efficiency less than 98 percent.
     For those existing process vent control devices that are
achieving less than 98 percent, the EPA has provided emissions
averaging as an alternative compliance option.  An emissions
credit from control of another emission point in the facility
                              2-1

-------
can be used to offset the emission debit generated by the use
of a process vent control device with less than 98 percent
efficiency.  For small production facilities,  the magnitude of
the emissions debit generated by controlling process vents to
efficiencies between 95 and 98 percent should be small.
Therefore, emissions averaging is also a viable option for
these facilities.
     Comment;  One commenter (A-90-19:  IV-D-32) stated that
the EPA has not demonstrated that RCT achieves 98 percent
control for each HAP and that the 98 percent level of control
appears to be based only on the VOC removal levels used in
past NSPS.  One commenter (A-90-19:  IV-D-70)  supported the
use of thermal incineration for control of process vents and
other streams where reasonable.
     One commenter (A-90-19:  IV-D-107) agreed that the
proposed emission limits are achievable and should be
promulgated as part of the final rule.  Two commenters
(A-90-19:  IV-D-70; IV-D-99) stated that thermal incinerators
can provide control greater than 98 percent ORE if properly
operated, while another commenter  (A-90-19:  IV-D-90)
requested that a 99.9 percent DRE be required for all
combustion devices.
     Response;  The EPA would first like to reiterate that
control by thermal oxidation is not specifically required by
the HON process vents provisions.  Thermal oxidation is simply
the RCT whose performance level must be met by any controls
intended to comply with the HON process vents provisions.  The
commenter correctly states that 98 percent control is based on
studies used to determine VOC control levels for past NSPS and
has not been proven by testing for each individual HAP.  These
two issues do not weaken the EPA's decision for 98 percent
control of HAP's for the following reasons:  (1) nearly all
organic HAP's are also VOC; and (2) HON compliance is not
based upon control of each individual HAP.  Compliance with
the HON may be based upon either total organic HAP or TOG.
Clearly, a control device might have a higher level of control
for one particular HAP than for another, but compliance  is
                              2-2

-------
based on the overall reduction of total organic HAP or TOC
from an emission point.
     The 98 percent level of control was chosen because it has
been shown to be uniformly achievable by combustion devices.
As stated earlier, test data to demonstrate efficiency in a
thermal incinerator is not available for each individual HAP.
However, the efficiency conclusions for a thermal incinerator
(98 percent DRE or an outlet concentration of 20 ppmv) were
based on test data using the most difficult VOC compounds to
combust, which included several HAP's.  Therefore, it was
concluded that the 98-percent reduction can be achieved for
total HAP (memorandum from David Mascone, EPA/CMS, to Jack
Farmer, EPA/CPB, "Thermal Incinerator Performance for NSPS,
Addendum," July 1990, Docket Number IV-90-19:  IV-B-1).  The
EPA recognized that thermal incineration may achieve greater
than 98 percent DRE, but test data show that levels greater
than 98 percent may not be uniformly achievable.
     Comment;  Two commenters (A-90-19:  IV-D-90; IV-D-100)
claimed that by focusing technical analysis of controls on
combustion,  the EPA has limited the flexibility in applying
alternative control technologies that are cost effective,
including pollution prevention.
     Response:  The technical analyses performed focused on
control by combustion because combustion is considered to be
the most universally applicable control for process vent
emissions, and accurate information was available for
estimating costs and impacts.  This in no way limits the
application of alternative controls provided that the
alternative control achieves the same level of control as
combustion:   98 percent emission reduction or 20 ppmv HAP at
exit of device.  Pollution prevention is in fact encouraged by
the EPA since it recovers valuable resources while reducing
emissions. .From this perspective, pollution prevention may be
viewed in the same light as recovery devices such as
condensers or carbon adsorbers.  If a pollution prevention
technique is employed within a facility's control strategy, it
must alter the characteristics of a Group 1 stream such that
                              2-3

-------
the stream is then classified as Group 2.   If this can not be
achieved by pollution prevention alone t then either emissions
averaging or an add-on control -device must be used _ in addition
to the pollution prevention technique.
     Comment ;  One commenter (A-90-19:  IV-D-113) stated that
the requirements for demonstration of alternative control
technology are both burdensome and unnecessary if a technology
meets a specified treatment objective.  The commenter
(A-90-19:  IV-D-113) recommended that the EPA establish a
treatment threshold for halogen removal in the process vents
section of the RON, but not dictate a specific technology to
achieve it.
     Response;  The process vent provisions are expressed as a
HAP or TOC percent reduction or concentration limit, and do
not require use of a specific technology.   The control
technology must be demonstrated to achieve the 98 percent
reduction or a concentration below 20 ppmv during the
performance testing to ensure it can meet this level.  If the
commenter 's alternative technology is one where performance
testing at the inlet and outlet would be infeasible, but there
is evidence that it achieves equivalent control, then the
commenter can apply for approval of the alternative technology
under §63. 6 (g) of the General Provisions.   This process
includes Agency Review and a Federal Register notice.
     However, because the proposed standard for halogenated
streams in §63. 114 (c) was written as an equipment standard
(scrubber following a combustor) use of any other technology
would have required an application and Federal Register
notice.  Therefore, the final rule is being revised to use an
emission limit (percent reduction or mass limit) format
instead of requiring a scrubber.  This will allow use of
technologies other than a scrubber to meet the halogen limits.
               One commenter (A-90-19:  IV-D-50) proposed that
an "Alternate Means" provision be added to allow application
of control technology less stringent than MACT if the facility
can demonstrate that thermal oxidation is unsafe and that
                              2-4

-------
applicable alternative technology cannot achieve a level of
control equivalent to RCT.
     Response;  The EPA again points out that thermal
incineration is not exclusively required by the HON
regulation.  Thermal incineration provides the basis for the
technical analyses and for comparisons to other control
technologies.  Possible alternatives include use of recovery
devices to increase the TRE to greater than 1.0, or emissions
averaging.
     Comment;  One commenter (A-90-19:  IV-D-34) suggested
that the incinerator definition be modified so that it would
include regenerative incinerators.
     Response;  The purpose of the definition is to
distinguish between incinerators and boilers.  The
incineration definition will be modified to include the
following sentence provided by the commenter (A-90-19:
IV-D-34) at the end of the definition to allow the use of
regenerative incinerators:
     "The above energy recovery section limitation does
     not apply to an energy recovery section used solely
     to preheat the incoming vent stream or combustion
     air."
     Comment:  Two commenters (A-90-19:  IV-D-70; IV-D-99)
stated that because many boilers and process heaters employ
staged combustion, which may lower the overall temperature and
lengthen the flame zone of the combustion section, the vented
VOHAP stream should be introduced into the combustion flame
zone as near to the burner fuel inlet as possible while
maintaining good mixing.
     One commenter (A-90-19:  IV-D-86) considered the
definition of flame zone too restrictive and suggested that a
residence time requirement be established as an alternative.
     Response;  It was determined that as long as the process
vent stream passes through the flame zone, the temperature and
residence time achieve the required level of combustion
efficiency.  The EPA references "Reactor Processes in the
Synthetic Organic Chemical Manufacturing Industry - Background
Information for Promulgated Standards," EPA-450/3-90-016b,
                              2-5

-------
March 1993 to support the decision.  The definition of flame
zone in the regulation was written broadly enough to include
various types of boiler configurations .*_ As .proposed,  the
definition allows the stream to be introduced as near-to-the
burner fuel inlet as possible while maintaining good mixing.
However, in certain configurations, it may be too restrictive
to adopt these suggested requirements into the definition,
especially since another commenter thought that the proposed
definition of flame zone was too restrictive.  Furthermore,
such wording is very subjective and would therefore be
difficult to enforce.  The EPA determined that the requirement
that the vent stream be introduced into the flame zone was a
simpler requirement to follow than establishing and measuring
the residence time.  For these reasons, the definition of
flame zone remains unchanged.
     Comment;  One commenter (A-90-19:  IV-D-69) suggested
that a vent stream be allowed to be introduced with secondary
combustion air when required by compatibility or safety
reasons if compliance with the 98 percent ORE or 20 ppmv exit
concentration requirement is maintained.
     Response;  The proposed and final process vent provisions
permit a vent stream or streams to be introduced with
secondary combustion air if a facility so chooses.  For
boilers with heat input capacities less than 44 MW or if the
vent stream is introduced with the combustion air, a
performance test and more monitoring, reporting, and
recordkeeping are required than if the stream were introduced
as or with primary fuel.
     Comment;  Two commenters (A-90-19:  IV-D-32; IV-D-34  and
IV-G-4) requested that flares be allowed as the primary
control device for halogenated streams, if the aggregated
halogen content at the flame tip under routine operating
conditions would not exceed 4 Ib/hr as halogen atoms.  The
commenters (A-90-19:  IV-D-32; IV-D-34 and IV-G-4) pointed out
that such a change would be consistent with some existing
State regulations.  For example, one commenter  (A-90-19:
IV-D-32) said TACB Standard Exemption 80 exempts from new
                              2-6

-------
 source review up to 0.45 Kg/yr  (1.0 Ib/hr) of HC1 emissions.
 Another commenter  (A-90-19:  IV-D-88) cited a RCRA limitation
 on hazardous waste incinerators in proposing that a 4 Ib/hr
 halogen emission limit be.applicable to any combustion device.
 Another commenter  (A-90-19:  IV-D-32) said that existing
 controls in similar facilities were not considered in
 selecting the halogen cutoff level.  One commenter
 (A-90-19:  IV-D-34 and IV-G-4) stated that a flare is the only
 practically applicable control device for streams with highly
 variable flow and heat content, regardless of halogen content.
     One commenter (A-90-19:  IV-D-113) cautioned that
 experience with scrubbing and other halogen controls has
 demonstrated that the required 99 percent removal efficiency
 is extremely prescriptive for many halogenated process vents,
 especially those at or near the 200 ppmv threshold and those
 containing elemental bromine or chlorine  (Br2 or Cl2)•  One
 commenter (A-90-19:  IV-G-4) stated that Texas regulations
 require halogen control devices with 95-percent removal
 efficiency.  One commenter  (A-90-19:  IV-D-113) further stated
 that the 0.5 mg/Nm3 (mg/scm) halogen emission limit is ten
 times more stringent than the most stringent requirement they
 are aware of.  The commenter (A-90-19:  IV-D-113) proposed
 that the emissions limit be revised to 5 mg/Nm3  (mg/scm) to
 provide a limit that is technically achievable and provide
 consistency between European and U. S. standards so as to not
 hinder competitiveness.  One commenter (A-90-19:  IV-D-88)
 questioned the validity of the 0.5 mg/scm halogen emission
 limit because it was derived from the halogen detection limit
 rather than on a consideration of the characteristics of waste
 streams.
     Response;  The EPA agrees with the commenters that a mass
 limit for defining halogen streams will provide greater
 flexibility for compliance without reducing the stringency of
 the rule.  Based on an analysis of scrubber performance
 reported in the ethylene dichlorine questionnaire responses
 supplied by SOCMI facilities, 11 of the 12 scrubbers were
achieving a 99-percent reduction of hydrogen chloride or a
                              2-7

-------
total halogen mass flow rate below 0.45 kg/hr (1.0 Ib/hr).
The median mass emission rate exiting those scrubbers was
about 0.45 kg/hr of total HC1 and Cl2«  Based on the available
data, it is not demonstrated that values lower ..than .that would
be uniformly achievable.  The commenter (A-90-19:  IV-D-32)
said TACB has a 0.45 kg/hr exemption.  Therefore, the rule
will be revised to define a halogen stream as a stream
containing 0.45 kg/hr or greater of halogen atoms and require
a 99-percent reduction of total halogen atoms or reduce the
halogen emissions to less than 0.45 kg/hr.
     However, an allowance will be made for existing
scrubbers.  As indicated by the questionnaire responses and
comments, there are some SOCMI units that currently have
scrubbers that achieve between 95- and 99-percent reduction
and that would not achieve a 99-percent reduction or
0.45 kg/hr emission rate.  The EPA's national cost estimate
did not include costs to replace existing scrubbers.  The
emission reduction obtained from replacing a 95-percent
efficient scrubber with a 99-percent efficient scrubber would
be small.  Therefore, the final rule will allow sources that
had halogen control devices as of proposal of the HON to
achieve 95-percent reduction or an emission rate below
0.45 kg/hr.
     Comment;  One commenter (A-90-19:  IV-D-88)  questioned
the appropriateness of requiring a scrubber on process vents
with halogen emissions less than some combustion sources.  The
commenter (A-90-19:  IV-D-88) stated that a coal-fired boiler
exhaust may contain 70 ppmv HCl or more.
     Response;  The EPA's decision to require a combustor and
scrubber on a process vent stream that has halogen emissions
less than some combustion sources was based on the results of
analyses of cost effectiveness of the combustor/scrubber
combination.  The TRE index value is a measure of cost
effectiveness of control and the TRE calculation for
halogenated streams is based on application of a combustor
followed by a scrubber.  Halogenated process vent streams are
required to be controlled only if they have TRE index values
                              2-8

-------
less than or equal to 1.0.  For Group 1 streams  (those with
TRE < 1.0), application of a combustor and-a scrubber is	
reasonable.  The halogenated stream definition has been
revised based on a mass emission rate, as-described in -----
previous responses.
     Comment;  One commenter (A-90-19:  IV-D-34  and IV-G-4)
suggested that provisions be made for flares to  be used as a
backup control device for halogenated streams while the
primary control device is not operating, such as process
start-up and shutdown and primary control device malfunction
or maintenance.
     Response;  In the proposal regulation, an owner or
operator is not allowed to use a flare as a primary control
device on a halogenated vent stream.  However, as provided in
the proposed General Provisions found in subpart A, it may be
possible to flare such a stream as an alternate, back-up
control in case of start-up, shutdown, and malfunction of the
primary control device.  In order to gain approval for the use
of flaring as an alternate control during a start-up,
shutdown, and malfunction episode, the owner or  operator would
need to submit a start-up, shutdown, and malfunction plan as
described in §63.6(e)(3).  Each plan would be reviewed and
approved on a site-specific basis to determine if it is an
appropriate back-up plan in case of start-up, shutdown, and
malfunction of the primary control device.
     Comment;   Two commenters (A-90-19:  IV-D-70; IV-D-99)
stated that flare operational standards and parameters and
parameter monitoring should conform to 40 CFR 60.18.
     Response;  The EPA agrees with the commenter.  The flare
operational standards in 40 CFR 63.11(b) are the same as those
in 40 CFR 60.18.
     Comment;   Two commenters (A-90-19:  IV-D-70; IV-D-99)
suggested that scrubbers used to control halogenated emission
streams use either a continuous purge of the scrubbing fluid
or maintain a two to ten day supply of scrubbing fluid on
site.   Additionally,  the commenters (A-90-19:  IV-D-70;
IV-D-99)  recommended that an extra scrubbing fluid pump and
                              2-9

-------
spare parts be required to be on hand at the site or available
within one working day in order to minimize^ scrubber- downtime,
     Response:  The EPA understands- the. commenters^ .concerns
over control device downtime; however, the EPA has" chosen not
to include specific requirements in the HON such as those
listed by the commenters.  The potential for assessing non-
compliance fees should provide sufficient incentive for owners
or operators to keep scrubbers operating properly with timely
maintenance.
     Comment;  One commenter (A-90-19:  IV-D-77) suggested
that a water scrubber should be allowed as a process vent
control if it can be proven that the scrubber removes
98 percent of the organic HAP.  The scrubber underflow would
be sent to the plants wastewater treatment facility.
     Response;  This type of control is allowed under the HON
process vents provisions, provided that the device
demonstrates 98 percent control through performance testing,
and it is not used as a product recovery device.   (Product
recovery devices are considered part of the chemical
manufacturing process and can be used to achieve a TRE greater
than 1.0, but cannot be used to comply with the 98 percent
emission reduction provision.)
     Comment:  Four commenters (A-90-19:  IV-D-70; IV-D-85;
IV-D-99; IV-G-7) presented concerns about the inability of
condensers and carbon adsorbers to meet RCT for process vents.
Possible problems mentioned by the commenters (A-90-19:
IV-D-70; IV-D-99) occur when condenser coils freeze up and
when a mixture of compounds is controlled by a carbon
adsorber.
     Response;  The EPA is aware of the fact that condensers
and carbon adsorbers may not meet the 98 percent
reduction/20 ppmv level of control required by the process
vent provisions; however, for process vents, if these devices
are used for product recovery, they are not considered to be
control devices, and cannot be used to meet 98 percent
reduction/20 ppmv provisions.  As recovery devices, they can
be used to maintain a level of performance such that the
                              2-10

-------
outlet stream from the device has a TRE greater, than 1.0.  A
condenser or carbon adsorber applied to a storage vent or
transfer rack must meet the RCT requirements for those source
types, 95 percent or -9:9 ^percent control,- respectively^. .- .In any
case, if the device is used to comply with the 98 percent
reduction or 20 ppmv level of control, the level of control of
these devices must be proven through a performance test.
     Comment;  Three commenters (A-90-19:  IV-D-70; IV-D-85;
IV-D-99) also expressed concern over the use of ozone
depleting chemicals in condensers applied as controls.
     Response:  The EPA is providing neither an incentive nor
disincentive for the use of ozone depleting chemicals in
condensers.  The use of those chemicals is addressed under
Title IV of the Act.
     Comment;  One commenter (A-90-19:  IV-D-107) claimed that
catalytic incineration is a proven, effective control
technology already in use within the SOCMI as an effective VOC
control and will prove to be an efficient HAP control as well;
however, two commenters (A-90-19:  IV-D-70; IV-D-99) pointed
to problems with catalyst poisoning when catalytic
incineration was used (see section 2.4.1).
     Response:  To meet the requirements of the regulation,
any type of control device, including a catalytic incinerator,
can be used as long as it reduces HAP emissions by 98 weight-
percent or to a concentration of 20 ppmv, on a dry basis,
corrected to 3 percent oxygen, whichever is less stringent.
To ensure proper operation of the selected control device, the
regulation requires that ranges for certain parameters must be
established in a performance test and then monitored
periodically.  When a catalytic incinerator is used to achieve
compliance, a daily average temperature difference across the
catalyst bed must be established based on the performance test
and other documentation and reported in the NCS or established
in the operating permit.  Once the temperature difference is
established, it must be measured continuously, and a daily
average must be calculated each day.  If the catalyst bed
becomes poisoned, the temperature difference measured across
                             2-11

-------
the catalyst bed would Likely be outside the range established
in the NCS or the operating permit, indicating a problem.
2.2  IMPACTS ANALYSIS
2.2.1  Cost Analysis - :. - - =
     Comment ;  Two commenters (A-90-19:  IV-D-58; IV-D-62)
argued that the incremental emissions reduction gained by
exceeding the MACT floor is unjustified and that if the EPA
regulates at a stringency above the floor requirements, the
EPA must show that the cost of exceeding the floor
requirements are outweighed by substantial emission reduction
benefits.  One commenter  (A-90-19:  IV-D-58) said the
incremental emissions reduction of 1.3 percent from the floor
to the selected option is likely to be in the realm of a
statistical aberration given emission estimation
uncertainties .
     Response;  The average characteristics of the floor was
equivalent to about $3,000/Mg of organic HAP reduced.  The EPA
believes the estimates of emission reduction and costs are
sufficiently accurate for use in regulation development, and
the commenter did not provide specific details regarding their
concern about estimation uncertainty.  The EPA further
believes that the incremental cost associated with the
incremental emissions reduction is acceptable and justified
and brings the HON level of stringency more in line with the
cost-effectiveness in previous NSPS and the CTG.
               One commenter (A-90-19:  IV-D-32) stated that
the EPA's cost analysis was generally correct but contended
that there were two exceptions.  First, the commenter
(A-90-19:  IV-D-32) contended that some of the EPA's
assumptions were too conservative and lacked substantiation.
Second, the commenter (A-90-19:  IV-D-32) claimed that
baseline controls for nonattainment areas were applied to
facilities that are not in nonattainment areas.
     Response:  The EPA thanks the commenter for the general
support of the EPA's cost analysis.  Although the commenter
disagreed with some of the EPA's assumptions, no specifics
were given concerning which assumptions were being questioned
                              2-12

-------
and no alternatives were suggested by_the_commenter.
Secondly, the commenter_incorrectly stated that controls
required for facilities in non-attainment areas were
considered when evaluating baseline control^ level..  Only
existing State regulations were included in the baseline
control analysis.
     qomment;  Two conunenters (A-90-19:  IV-D-32; and
IV-D-112) stated that the HON did not consider the costs of
upgrading existing controls to 98 percent ORE performance and
presented an example of a catalytic incinerator (95 percent
ORE) installed for compliance with the Air Oxidation NSPS.
One commenter (A-90-19:  IV-D-32) also suggested that the TRE
measurements be allowed at the exit of a control device
achieving at least 95 percent control efficiency.
     Response;  The Benzene NESHAP, vinyl chloride NESHAP, and
regulations in five states were reviewed.  All of these
regulations required 98-percent control, except for one State
regulation that requires 85-percent control.  Therefore, all
vents included in the HON baseline were assumed to be
controlled to 98 percent except vents in the State requiring
85-percent reduction.  These latter vents were assumed to be
controlled with condensers, and costs were estimated for
upgrading to incinerators.  Thus, the HON does consider the
costs of upgrading existing controls to 98-percent ORE
effectiveness.
     The Air Oxidation NSPS, which requires 98-percent
reduction of VOC emissions from new air oxidation process
vents, was not considered for the HON baseline, which means
any vents that would be "caught" by this NSPS were assumed to
be uncontrolled in the HON baseline analysis.  Incorporating
controls under the Air Oxidation NSPS, or any additional rule
in the HON baseline analysis, would cause the HON control cost
estimates to be lower.  The HON costs of control are
overestimated.
     qofflfflfflTt;  Two conunenters (A-90-19:  IV-D-69; IV-D-75)
stated that the EPA erroneously assumed that all process vent
streams are centrally collected and routed to a single control
                             2-13

-------
device.  One commenter (A-90-19:  IV-D-32)  gave the following
reasons why vent streams may not be collected together:  mixed
streams may create serious safety concerns, vent proximity or
energy requirements may make .: central, collection . impractical,
and various streams may be incompatible with the design of a
single control device.  Two other commenters (A-90-19:
IV-D-79; IV-D-97) also pointed out the potential safety
hazards of plant -wide manifolding of vent streams.  One
commenter (A-90-19:  IV-D-86) supported the determination of
MACT control device cost effectiveness on a dedicated vent
basis.  The commenter (A-90-19:  IV-D-86) acknowledged that
vents may at times be combined cost effectively and safely,
but stated that it would be inappropriate to assume all small
vents may be manifolded together.
     Response;  The EPA believes the commenters have
misinterpreted the cost analysis.  For estimating cost of
control, the EPA assumed that reactor vents from the same CMPU
were combined, air oxidation vents from the same CMPU were
combined, and distillation vents from the same CMPU were
combined.  The EPA did not combine vent streams from different
CMPU's, nor did the EPA combine vent streams from different
vent types (air oxidation, distillation, reactor) .  The EPA
did assume that reactor vents from the same CMPU would be
located near each other, and no information was received to
discredit this assumption.  Therefore, the EPA does not
believe that safety, vent proximity, and stream
incompatibility are of concern because of the methodology used
for the national impacts analysis.
               One commenter (A-90-19:  IV-D-68)  stated that
the purchase of computers, running of conduit, installation of
wiring, programming, and all auxiliary equipment required for
monitoring will raise costs of control to TRE's of 10.0 or
greater, yet these costs do not appear to be included in the
TRE calculation or cost model.
     Response:  The burden for an individual vent should be
small since many plants already have process control computer
systems or would purchase and program computer systems, because
                              2-14

-------
 of the need to monitor other emission points.   Most  vent
 parameter monitors (e.g.,  temperature monitors)  are  very
 inexpensive (relative to the control  device  costs  included in
 developing the TRE equations).   Therefore,  including these
 costs would not be expected to  significantly increase the TRE
 index values.
      Also,  the standards do not require use  of  computers.
 Provisions have been added to §63.151(f)  allowing  sources to
 request alternative monitoring  for non-automated systems.
 Costs for computerized recordkeeping  and reporting were
 calculated as  part of the national burden estimate and were
 considered in  developing the standards.
 2.2.2  Emission Estimates
      Comment;   One commenter (A-90-19:   IV-D-71) disagreed
 with the assumption that uncontrolled emissions are  linearly
 related to the production capacity of a give production
 process.
      Response;   A linear relationship was considered to be the
 best estimate  based on the available  data.   Although a linear
 estimate may not be precise for a given production process,
 the EPA regards these estimates as a  reasonable representation
 of emissions on a nationwide basis.   The commenter did not
 provide data relating production capacity and uncontrolled
'emissions on which to base a revision in the emission
 estimation methodology.
 2.3  APPLICABILITY AND GROUP I/GROUP  2 DETERMINATION
 2.3.1  Applicability
      rmrnflent;   one commenter (A-90-19:   IV-D-92) asked how the
 TRE applies to particulate HAP's.
      Response;   The EPA does not expect organic HAP
 particulate emissions from the  regulated processes.   The  TRE
 would not apply if such emissions should occur.
      C^flffRfrlrt;   Two commenters (A-90-19:   IV-D-73;  IV-D-113)
 requested that the standard be  clarified by  expressing all
 applicability  and treatment criteria  on a ppmv  basis,
 specifically changing the 50 ppmw organic HAP concentration in
 the process vents definition.   Another commenter (A-90-19:
                              2-15

-------
IV-D-71) asked that the 0.005 weight percent exemption
included in the process vent definition be expressed in ppmv
instead of in ppmw.  One -commenter (A-90-19: -IV-D-77X
recommended that the 50 ppmv Group 2 criteria be used in lieu
of the 0.005 weight percent (50 ppmw).
     Response;  The ppmw unit was used to express the
applicability criteria of 0.005 ppmw because the data on which
the decision was based were expressed as ppmw.   The ppmw unit
has been retained in the final HON in order to avoid using an
arbitrary conversion from ppmw to ppmv.
     Comment;  One commenter (A-90-19:  IV-D-86) pointed out
that under certain circumstances a vent stream could be both
greater than 0.005 percent HAP by weight and less than 50 ppm
HAP by volume.  The commenter (A-90-19:  IV-D-86) thought the
proposed HON would in this case be unclear as to whether or
not the vent stream is subject to control, and recommended
adding a 0.002 percent by volume clause to the process vent
definition.
     Response;  The commenter is correct in stating that a
stream could be both greater than 0.005 weight percent and
less than 50 ppmv.  In this case, the stream would be a
Group 2 process vent.  The following protocol must be followed
in making the above determination:  (1) does the HON apply to
the stream; and (2) is the stream Group 1 or Group 2.  If the
stream contains less than 0.005 HAP weight percent, the stream
is not considered a process vent, the HON is not applicable,
and no further determination is required.  Assuming the stream
meets the weight percent applicability criteria, a group
determination must be made.  This may be done using the TRE
equation, the low-flow level criterion, and/or the low
concentration level criterion.  The commenter's example uses
the low concentration criterion of 50 ppmv HAP.  If the stream
in question is below 50 ppmv HAP, it is classified as a
Group 2 stream and must comply with the Group 2 process vent
requirements.  If the stream is equal to or greater than
50 ppmv HAP and has a flow greater than 0.005 scmm and a TRE
                              2-16

-------
less than or equal to 1.0, it would have to comply with the
Group l process vent requirements.
     Comment:  Three commenters  (A-90-19:  IV-D-32; IV-D-98;
IV-D-112) favored including a de minimis flow rate of
0.005 scmm in the subpart F process vents definition instead
of using the 0.005 scmm flow rate to identify Group 2 vents
without TRE calculations.
     Response;  A flowrate of 0.005 scmm is given in the
regulation to distinguish Group 2 process vent streams that
are not required to perform a TRE calculation.  However, it
was not the EPA's intent to exempt those Group 2 process vent
streams with a flowrate of 0.005 scmm from all requirements of
the HON.  The EPA's intent was to exempt such streams from the
Group l control requirements, but to require minimal reporting
and recordkeeping necessary to verify that the process vent
was correctly classified as Group 2 and to require reporting
of flow rate changes that cause such a vent to become Group 1.
In order to accomplish this intent, the process vent stream
definition does not incorporate a flowrate criterion.
     Comment;  One commenter (A-90-19:  IV-D-77) noted that
the weight percent applicability and low concentration Group 2
criteria seem to indicate the criteria are based on individual
HAP's while the testing methods give results in total HAP
concentration.  The commenter (A-90-19:  IV-D-77) asked that
the methodology be changed to match the definitions, or vice
versa.
     Response;  The EPA has revised the definition of process
vent in S63.101 and Group 2 process vent in §63.111 to clarify
that the applicability criteria are based on total organic
HAP.
     Cr>TfflTC«an.^•  Two commenters (A-90-19:  IV-D-32; IV-D-69)
asked that a halogen mass flow rate be included in the
definition of a halogenated stream to allow certain 200 ppmv
and higher halogen vent streams with low mass flow rates to be
vented to a flare (i.e., a control device without an acid
scrubber).
                              2-17

-------
     Response;  The EPA agrees with the commenters and the
definition of a halogenated stream has been revised to include
a halogen atom mass flow rate of 1.0 Ib/hr or greater instead
of a concentration basis.
     Comment;  One commenter (A-90-19:  IV-D-85) stated that
the EPA offers no justification for exempting vents associated
with wastewater treatment from the requirements for process
vents.
     Response:  Vents associated with wastewater treatment are
exempt from the process vents requirements, but not exempt
from control.  Control requirements for vents associated with
wastewater treatment are in §63.139.
     Comment;  One commenter (A-90-19:  IV-D-112) claimed that
it is not economically or environmentally feasible to control
vents with insignificant emissions.
     Response;  The commenter has used some subjective and
vague terms such as "insignificant emissions" and
"environmentally feasible" without supporting data for
clarification.  With regard to "insignificant" emissions, the
EPA assumes that the commenter supports the establishment of
mass emission rates below which controls would not be
required.  The economic feasibility of controlling a vent
stream is determined by the TRE calculation.  The EPA has
attempted to identify streams with high or "unreasonable"
cost-effectiveness through the establishment of a
Group 1/Group 2 classification based either on TRE or on low
flow and low concentration levels.  The emissions from streams
qualifying as Group 2 under these criteria would likely be
considered "insignificant" by the commenter.
2.3.2  Group 1/Group 2 Determination
     (;9Pfflent;  Two commenters (A-90-19:  IV-D-86; IV-D-92)
said that since the TRE calculation is based on individual
vents, it would be inappropriate to apply the TRE
determination to combined vents.
     Response:  The EPA agrees that the TRE should be applied
on an individual vent basis as the proposed rule states.  The
point of measurement is the outlet of the  final product
                             2-18

-------
recovery device  (if any recovery device is present) and prior
to any subsequent combination or release to the atmosphere.
In cases where vents are already grouped to a common header,
compliance may be achieved through the application of a
combustion device in order to avoid the TRE calculation
altogether; or the TRE's of individual streams may be
determined, and  if any are Group I, these can be controlled or
included in an emissions average.
     Comment;  One commenter (A-90-19:  IV-G-4) requested that
the proposed HON be revised so that mixing of streams prior to
a product recovery device not be considered dilution and that
TRE determination be performed after the last product device.
The commenter (A-90-19:  IV-G-4) noted that compliance for new
sources desiring to use product recovery on a number of
streams would be very costly if duplicate product recovery
devices would be required.
     Response;  The EPA intended for the measurement for TRE
determination to be taken following the final product recovery
device.  Prior to this final product recovery device, mixing
of streams is allowed; however, once the process vent stream
passes through the final product recovery device, the
measurement for TRE determination must be taken" prior to any
further mixing of streams.  The final rule has been revised to
clarify this requirement.
     Comment;  One commenter (A-90-19:  IV-D-92) asked for
clarification as to whether individual streams from
distillation columns are measured or whether group vents from
the "process unit" are measured.  The commenter
(A-90-19:  IV-D-92) added that this is also a point of
ambiguity in NSPS, subpart NNN.
     Response;  The point at which all testing must be done
for the purpose of group determination is after the final
recovery device and prior to mixing with any other stream or
streams.  Therefore, the individual streams from each recovery
device would be tested if testing is necessary.
     comment-;  one commenter (A-90-19:  IV-D-62) reasoned that
since the ORE for boilers and process heaters is already
                             2-19

-------
proven to meet or exceed 98 percent, there is no need for TRE
determination for vents routed to fuel gas systems where the
fuel gas is used as the primary fuel and the process vent
definition should be rewritten to exclude these streams.
     Response;  A TRE determination is not necessary for vent
streams complying with the 98 percent reduction requirements,
regardless of the control device used.  However, such process
vent streams are still considered process vents, and are
subject to monitoring, reporting, and recordkeeping
requirements in the rule.  A performance test is not required
for boilers/process heaters with a heat input greater than
44 MW or where the vent stream is combined with the primary
fuel.
     Comment:  Two commenters (A-90-19:  IV-D-64; IV-D-73)
supported the exclusion for recalculating TRE for changes that
are within the range on which the original TRE calculation was
based.  Another commenter  (A-90-19:  IV-D-34) requested that
the same exclusions be allowed for any changes included in an
operating permit or permit application.
     Response;  Any time that a change is made such that the
value of any parameter used in the TRE equation for a process
vent stream (e.g., flow, organic HAP emissions, TOG emissions,
or heating value) is outside the range on which the original
TRE calculation in the NCS was based, the TRE index value must
be recalculated.  If the change is within the range used to
determine the original TRE, then the TRE does not need to be
recalculated.  If a parameter is within the range included in
the operating permit but outside of the range used as the
basis for the TRE determination, the TRE value would still
need to be recalculated.
     As stated in $63.4 of the General Provisions, an owner or
operator who is subject to an emission standard would comply
with the requirements of the emission standard regardless of
whether:  (1) an operating permit had been issued to that
source, or (2) the operating permit has been revised to
include the emission standard requirements.  In most all
cases, the requirements given in the HON would override the
                              2-20

-------
requirements given in the operating permit.  An exception
would be where the operating permit contains more stringent
requirements than those included in HON.
     Comment;  One commenter (A-90-19:  IV-D-34) objected to
adjusting ©2 concentration to 3 percent for the purpose of
determining Group I/Group 2 status via the low HAP
concentration exemption.  Another commenter (A-90-19:
IV-D-70) added that the correction is inappropriate in cases
where very little 02 is present, such as nitrogen blanketing.
     Response:  The commenters1 objection is well-founded.
The correction to 3 percent 02 in S63.115(c) was inappropriate
for determining Group I/Group 2 status and has been removed
from the final rule.
     Comment;  Four commenters (A-90-19:  IV-D-52; IV-D-79;
IV-D-86; IV-D-97) requested that testing for the purpose of
determining group status by the TRE calculation be allowed
after any existing control devices (if any controls are
present) for vents which have applied controls to existing
sources.
     Response;  If the device is the final recovery device,
TRE testing is to be performed after the final recovery device
(i.e., at the outlet prior to release to the atmosphere or
prior to a combustion device) and prior to mixing with any
other streams.  If the control device is an existing
combustion device, no TRE determination is required provided
the combustion device is achieving a 98 percent level of
control.  If an existing control device does not achieve
98 percent HAP reduction or 20 ppmv, then a group
determination must be made for the stream prior to the
combustion device.  If this is not done, a situation could
exist such that a stream is Group 1 at the inlet and Group 2
at the outlet of a less than 98 percent efficient combustion
device.  If group status were determined at the outlet of the
device, it would be allowed to continue to operate at a level
of performance less than MACT.  This would be contrary to the
intent of the HON regulation.
                             2-21

-------
     In the case of an existing combustion device with less
than 98 percent efficiency, the following options are
available:  (1) alter the process or apply a recovery device
so that the stream is Group 2 prior to the combustion device;
(2) upgrade the existing device so that it achieves 98 percent
HAP or TOC reduction or 20 ppmv outlet concentration;
(3) replace the existing combustion device with a new one that
achieves 98 percent efficiency; or (4) utilize an emissions
averaging plan so that the emissions debits from the
underperforming devices are compensated for by credits
elsewhere in the source.
     Comment;   Several commenters (A-90-19:  IV-D-48; IV-D-92;
IV-D-112; IV-D-113) supported the use of engineering estimates
for calculating TRE where TRE is greater than 4.0.  One
commenter (A-90-19:  IV-D-32) said that within the expected
accuracy of engineering estimates, their analysis indicates
that a TRE cutoff value of 3.0 will capture all Group 1 vents
in the EPA BID draft with the exception of a single high flow,
low HAP concentration vent.  The commenter (A-90-19:  IV-D-32)
recommended that a TRE of 3.0 be used to establish the
calculation-based cutoff to provide relief to limited testing
resources from the TRE testing determination procedure for
those process vents that are obviously Group 2.  The commenter
(A-90-19:  IV-D-32) suggested that an alternative cutoff value
be developed to exclude high flow, low HAP concentration vents
from the calculation-based alternative, but did not suggest an
alternative cutoff value.  Another commenter (A-90-19:
IV-D-86) supported TRE testing when the calculated TRE is less
than 4.0.
     Response;  The TRE value of 4.0 has not been changed
since the EPA considers this value reasonable.  Engineering
judgement is allowed in determining group status provided that
the TRE of the stream is calculated and shown to be greater
than 4.0.  If the TRE calculation results in a TRE less than
or equal to 4.0, the TRE inputs must be measured and the TRE
recalculated,  or the flow or concentration must be tested to
qualify for Group 2 status.  This does not penalize the
                              2-22

-------
facility, but in fact allows the facility to avoid full TRE
testing that would otherwise be required for a Group 2 process
vent with TRE greater than 1.0 and less than or equal to 4.0.
Additional calculation-based cutoffs for. high flow, low HAP
concentration streams would complicate the rule by requiring
additional calculations or testing to determine which process
vent streams are high flow, low HAP concentration.
     Comment :  Several commenters (A-90-19:  IV-D-48; IV-D-58;
IV-D-64; IV-D-73; IV-D-78) proposed that engineering
calculations or operational data be used for the process vent
flow rate and process vent HAP concentration Group I/Group 2
determination.  One of the commenters (A-90-19:  IV-D-64)
suggested that it would then be up to the discretion of the
regulatory agency to judge the technical accuracy of the data
used in the calculation.
     Response;  If only flowrate or concentration is selected
for process vent group determinations, testing is required.
Engineering judgement is allowed in determining the TRE index
value because a margin for inaccuracies in estimation has been
included.  If the TRE index value is less than 4.0, testing is
required to ensure the accuracy of the TRE index value.
Engineering judgement is allowed for flowrate and
concentration estimates used in the calculation.
               Three commenters (A-90-19:  IV-D-32; IV-D-33;
IV-D-112) suggested that engineering judgement or process
knowledge be allowed for determining the classification of a
halogenated vent stream.  Commenters (A-90-19:  IV-D-32;
IV-D-112) stated that this could be used specifically when
halogen status made no difference in Group I/Group 2
classification via the TRE calculation.
     Response ;  Although the proposed rule did not contain the
explicit language to allow the commenter's suggestion, it was,
in fact, the EPA's intent to permit engineering judgement for
halogenated streams when halogen status does not affect the
outcome of the group status determination and the calculated
TRE is greater than 4.0.  The final rule has been revised to
allow the engineering judgement in these cases.  When the
                             2-23

-------
calculated TRE is less than 4.0,  testing is required to more
accurately establish the TRE value.  ---_-__
               One commenter (A-90-19:   IV-D-74)  said that the
regulatory scheme in the proposed rule requires -that- the owner
estimate or measure emissions from every uncontrolled vent on
the plant site, and periodically recheck the estimate to
ensure its continued accuracy.  The commenter (A-90-19:
IV-D-74) then referred to 57 FR 62615 (§63.112).
     Response;  The EPA maintains the position that no owner
or operator would calculate emission estimates for every
emission point at the source in order to comply with the HON
and that the allowable emission level in §63.112  is one way of
expressing the standard.  The owner may elect to comply with
the RCT by adding a control device to each Group 1 vent and
not calculate emission estimates for those process vents.  The
owner may also measure only the process vent stream flowrate
to determine group status.  If the owner uses emissions
averaging, emissions only need to be calculated in debits and
credits for those points included in the average.
     Comment ;  One commenter  (A-90-19:  IV-D-64)  suggested
that the halogenated vent stream definition be revised so that
streams that are scrubbed to reduce halogens to less than
200 ppmv prior to a control device would not be considered
halogenated vent streams.
     Response;  The EPA agrees with the recommendation made by
the commenter that the halogenated vent stream definition and
requirements should be revised so that a scrubber to reduce
halogens prior to a control device can be used to comply.  The
scrubber used in this case does not need to be a product
recovery device.  In the final rule, a halogenated stream is
defined as having a mass emission rate of 0.45 kg/hr
(1.0 Ib/hr) or greater instead of a 200 ppmv concentration.
Streams with halogen emissions above 0.45 kg/hr must reduce
halogen emissions by 99 percent or to a level of 0.45 kg/hr.
A scrubber prior to a combustor could be used to meet this
emission reduction requirement.
                              2-24

-------
     rrnmiient:  Two commenters  (A-90-19:  IV-D-32; IV-D-88)
stated that the EPA has not provided a sufficient basis for
the establishment of the 200 ppmv halogen concentration
threshold in the definition of a halogenated vent stream.  One
commenter (A-90-19:  IV-D-32) advocated the following:
(1) establishment of a halogen mass emission rate (4 Ib/hr)
for the halogenated vent stream definition, (2) allowing
combustion by flare or incinerator of halogenated streams
below the suggested emission rate.
     Response;  The EPA reviewed scrubber efficiency data
provided in the ethylene dichloride questionnaire responses.
Based on the scrubber outlet halogen emissions data, the EPA
has established a 0.45 kg/hr (1.0 Ib/hr) or greater mass
emission rate as the definition of a halogenated stream.
Therefore, a stream containing 0.45 kg/hr or greater of
halogen atoms is considered halogenated and if combusted, must
reduce the halogen atom content by 99 percent or below
0.45 kg/hr of halogen atoms.  If a vent stream contains less
than 0.45 kg/hr of halogen atoms, this vent stream is not
considered halogenated and may be flared.
     Comment;  One commenter (A-90-19:  IV-G-4) suggested that
Group 1 and Group 2 status for existing halogenated streams
that are currently collected in a flare or fuel gas header for
control purposes should be determined after mixing and before
the emission control device.
     Response;  Group determination is not necessary if the
vent stream is controlled in a manner meeting the Group 1
control requirements in the regulation.  Group status is
determined on an individual vent basis to prevent dilution
from a Group 2 vent stream mixing with a Group 1 vent stream
and the resulting mixed stream being measured as a Group 2
stream and not being controlled.  The rule states that any
process vent stream containing halogens must be tested
individually prior to any mixing to determine the
concentration of halogens.  If the stream contains less than
0.45 kg/hr (1.0 Ib/hr) of halogen atoms, then the stream is
not considered halogenated and would be subject to the
                             2-25

-------
requirements for nonhalogenated streams.  However,  if the
stream contains 0.45 kg/hr or greater of halogen atoms, then
the stream would be considered a halogenated stream; if the
owner or operator routed the stream through-an incinerator, a
scrubber or other control combination achieving the 99-percent
reduction or reducing the halogen emission rate to less than
0.45 kg/hr would be required.
     Comment;  One commenter (A-90-19:  IV-D-88) proposed a
cost/reasonableness test that takes into consideration the
impact of requirements due to a particular component of a
mixed stream.  The commenter's (A-90-19:  IV-D-88)  specific
concern is the requirement of an incinerator and scrubber for
a mixed stream of both halogenated and non-halogenated HAP's.
The commenter (A-90-19:  IV-D-88) reasoned that although the
halogenated HAP content is sufficient to define the entire
stream as halogenated, the stream would not be classified as
Group 1 when only the halogenated compounds were considered
alone.  However, the commenter (A-90-19:  IV-D-88)  also stated
that the overall characteristics of a stream should be
considered when determining the appropriate control for a vent
stream.  The commenter (A-90-19:  IV-D-88) also suggested that
the EPA provide opportunities for special relief from the rule
if the halogenated process vent stream definition is not
changed from 200 ppmv.
     Response;  All TRE calculations for group determination
are to be based upon TOG and total organic HAP, regardless of
halogenation status.  In order to determine the appropriate
TRE index value (which is an index of cost-effectiveness of
control), the total flow, total HAP emissions, TOG emissions,
net heating value, and halogen concentration of the stream
must be used in the calculation.  No revisions have been made
to this aspect of the regulation.  Controls are required only
for Group 1 emission points which have been shown to have a
TRE of 1.0 or less.  The definition of a halogenated process
vent stream is revised as having a mass emission rate  of
0.45 kg/hr or greater of halogen atoms.  Analyses indicate
that control of such emission points is reasonable, and
                              2-26

-------
therefore special relief as requested by the commenter
(A-90-19:  IV-D-88) from the rule is not applicable.
               Two commenters 4A-90-19:  IV-D-74; IV-D-86)
supported the determination of MACT control" device cost
effectiveness on a dedicated vent basis.  One commenter
(A-90-19:  IV-D-86) acknowledged that vents may at times be
combined cost effectively and safely, but stated that it would
be inappropriate to assume all small vents may be manifolded
together .
     Response;  The EPA agrees with the commenter and thanks
them for their support.
2.4  COMPLIANCE DEMONSTRATIONS
2.4.1  Performance Testing
     Comment ;  One commenter (A-90-19:  IV-D-113) stated that
Method 2 is neither necessary nor the most accurate method for
high flow gas streams, such as air oxidation vents.  The
commenter (A-90-19:  IV-D-113)  recommended that the EPA allow
air oxidation process flow rates to be measured through
established mass balances or other means which can be
demonstrated to provide accurate measures in place of the use
of Method 2, 2A, 2C, or 2D.  Another commenter
(A-90-19:  IV-D-77) also warned of difficulties of measuring
flows that are highly variable, low volume, or near ambient
pressure with Method 2 and urged the use of methods developed
and validated for the operating conditions more typical of
process vents.  The commenter (A-90-19:  IV-D-77) stated that
using Method 2 may induce system upsets, i.e., non-steady
state flow.
     Response;  The EPA agrees with the commenter that
Method 2 is less likely to give a good average if the process
vent stream is highly variable.  However, other alternatives
for determining stream characteristics are available in the
HON besides Me.thod 2.  First, if an owner or operator can
document estimates of the flowrate and concentration of the
process vent stream derived from engineering assessment
(including process knowledge) ,  and the resulting TRE is
greater than 4.0, then the owner or operator does not need to
                             2-27

-------
use Method 2 to measure the process vent stream flow rate.
Note that the owner or operatorrwould be required to test the -
process vent stream_if the resulting TRE is less than or equal
to 4.0.  However, most vent streams-with ~a TRE of 4.-0 or less -
will be relatively large vent streams from continuous
processes, which are not highly variable,  and Method 2 will be
applicable to most of these streams.   Furthermore, if the
owner or operator encounters difficulties when using Method 2
under certain conditions, such as measuring low flow or highly
variable flow streams, alternate methods for measuring may be
validated according to Method 301 of 40 CFR part 63,
appendix A.  Once validated, those methods could be used
instead of Method 2.
     In addition, any owner or operator who wishes to use an
alterative monitoring method other than those discussed above
can submit an application for alternative monitoring
requirements to the Administrator as detailed in §63.8 of
subpart A of the General Provisions.
     Comment:  Two commenters (A-90-19:  IV-D-34; IV-D-71)
disagreed with the use of the 02 concentration adjustment to
3 percent 02 for non-combustion control devices.
     Response;  The commenter is correct.  The° final rule has
been revised so that only combustion devices are required to
correct to 3 percent 02.
     Comment;  One commenter (A-90-19:  IV-D-33) recommended
that each facility's permit be allowed to include bypass
provisions under certain sets of circumstances that take into
account the pollutants emitted, potential off-site impacts,
and volume of emissions and that bypasses permitted in such
provisions not be considered excursions for compliance
purposes.
     Response;  The General Provisions allow for a
malfunction, start-up, shut-down plan and bypasses covered in
that plan would not be considered excursions.  Bypasses not
covered under the plan would likely be considered excursions
or violations.
                              2-28

-------
               One commenter (A-90-19:  IV-D-92) suggested
allowing a limit switch to be attached to a bypass valve
(i.e., vent gas bypass) attached to a computer monitor because.
it may be helpful in minimizing data gathering requirements.
     Response;  The rule does not require emissions monitoring
of bypass valves, but requires a flow indicator to ensure the
vent stream is routed to the control device and not bypassed
to the atmosphere.  Provisions for monitoring parameters are
provided in the rule.
     Comment;  One commenter (A-90-19:  IV-D-92) reasoned that
variable flows and concentrations to combustion devices cause
testing of those devices to be non-representative, and
suggested that engineering judgement be allowed as a
supplement for, or in lieu of, combustion device testing.
     Response;  The EPA acknowledges that varying flows and
concentrations exist in process vents.  It is up to the
facility to use engineering judgement in choosing the process
conditions under which the source testing will be conducted.
These conditions should be chosen so that they are
representative of typical process operations.  With regard to
combustion device testing, if the combustion device achieves
the required emission reductions under one set "of process
conditions judged to be representative of the process, then
the EPA is confident that the combustion chamber temperature
during which the test was performed is adequate and the
standard will be achieved under normal operation of the unit.
The combustion chamber temperature is sufficient for
monitoring since variations in flow will cause the chamber
temperature to vary.
     qqfllTO^n^•  one commenter (A-90-19:  IV-D-92) pointed out
that non-flare combustion devices must prove 98 percent ORE by
performance testing, while flares are assumed to be 98 percent
effective with no required testing performance test.  The
commenter (A-90-19:  IV-D-92) suggested that the 98 percent
ORE for flares could be a flare temperature of 1500 °c with a
residence time of 0.75 seconds as in NSPS subpart QQQ or a
                              2-29

-------
temperature of 1400 oc with a residence time of 0.5 seconds as
in NESHAP subpart V.
     Response;  Theproper flare operation guidelines are
presented in §63.11 of the General Provisions.  These
guidelines must be followed to maintain compliance with the
RON regulation.
     Comment;  Two commenters (A-90-19:  IV-D-70; IV-D-99)
supported the establishment of compliance parameters for
incinerators, boilers/process heaters, and scrubbers by
performance testing and advocated the determination of
compliance based on hourly averages calculated from data
collected every fifteen minutes.
     Response;  As discussed in the reporting and
recordkeeping section (see section 7), compliance parameters
are established during the performance test, although
engineering assessment can also be used in establishing
parameter ranges.  For purposes of complying with the HON,
daily averages are used.  The commenter gives no rationale for
selecting hourly averages over daily averages.  The EPA's
rationale for selecting daily averages is discussed in the
above mentioned section.
     Comment;  One commenter (A-90-19;  IV-D-70) stated that
the DRE of boilers and process heaters (regardless of heat
duty) should be required to be established through initial
performance testing.  Three other commenters  (A-90-19:
IV-D-35; IV-D-64; IV-D-99) supported the testing exemption for
boilers and process heaters with heat  input greater than
44 MW.
     Response;  The initial performance test exemption is
appropriate for a boiler or process heater with heat input
capacity of 44 MW (150 million Btu/hr) or greater in which all
process vent streams are introduced into the flame zone and
for all boilers or process heaters in  which the process vent
streams are introduced with or as the  primary fuel.  Emission
factor calculations (AP-42), submitted test results, and
temperature and residence time calculations indicate that the
                              2-30

-------
expected ORE for boilers and process heaters with heat input
capacities greater than 44 MW would be greater than    :
98 percent.  The EPA references "Reactor Processes in the
Synthetic Organic Chemical Manufacturing Industry - Background
Information for Promulgated Standards," EPA-450/3-90-016b,
March 1993 to support the decision.  When the vent stream
passes through the flame front it would, on average, be
combusted at higher temperatures and longer residence times
than if introduced with combustion air.  This information
indicates that a process vent stream would achieve combustion
efficiency greater than the required 98 percent level.  For
this reason, it is not necessary to establish the emission
reduction of these boilers and process heaters through initial
performance testing.
               Two commenters (A-90-19:  IV-D-70; IV-D-99)
recommended that ranges for the following monitoring
parameters be established during the initial performance test
for thermal incinerators, boilers, and process heaters:
firebox exit temperature, and CO and ©2 concentrations in the
outlet stack gas.
     Response;  Based on previous incinerator performance
studies, temperature and residence time are the key parameters
which influence performance.  During the performance test for
combustion devices, the temperature is monitored and a range
established.  Any fluctuation in process vent flow rate will
be reflected by a change in temperature.  Therefore, the EPA
determined that temperature alone is sufficient to monitor
compliance for combustion devices which require a performance
test.  Other monitoring parameters, such as CO or ©2 outlet
concentrations may be requested under the alternative
monitoring parameter requirements in $63.114 and §63.151 of
subpart G.
     It is not appropriate to include specific values in the
regulation for the combustion parameters mentioned by the
commenter.  These combustion parameters need to be established
on a site-specific basis during the performance test, because
they would not be applicable to every situation.  These
                             2-31

-------
parameters are highly variable from one process to another
depending on the constituents x>f-the vent stream.------  ----
     Combustion devices -which- do not -require^ a_ performance
test (such as boilers and process heaters wittr-a heat capacity
design greater than 44 megawatts and a vent stream that is
introduced with the combustion air or a vent stream introduced
as or with the primary fuel) also do not require monitoring of
the combustion device, because the temperature and residence
time of these devices exceed the levels needed to achieve at
least a 98 percent reduction.
     Comment:  Two commenters (A-90-19:  IV-D-22; IV-D-73)
asserted that units that have undergone performance testing
for NSPS should not need to be tested for the HON if there
have been no process changes since the compliance tests.  One
commenter (A-90-19:  IV-D-22) stated that NSPS compliance
testing is rigorous, costly, and requires the same EPA methods
as the HON.   Another commenter (A-90-19:  IV-D-35) suggested
that any incinerator, boiler, or process heater that has
obtained an operating permit  (such as a State air permit or a
RCRA permit) and has existing data to prove 98 percent ORE
should not be required to conduct a performance test.
     Response;  The EPA agrees with the commenters that it
would be unnecessarily burdensome to require the owner or
operator of a unit to repeat identical tests required for
other compliance purposes, such as NSPS or RCRA, if no process
changes had been made to the unit since the test was
performed.  For this reason, the regulation has been revised
so that it does not preclude the use of previously conducted
tests if those tests were performed using the same test method
and no process changes have been made to the unit in the
interim.
     commer^•  One commenter  (A-90-19:  IV-D-22) recommended
that text be added to §63.116(b)(2) to clarify that a boiler
or process heater need not be tested when a process vent
stream serves as the primary fuel.  Another commenter
(A-90-19:  IV-D-34) pointed out that table 3 of
S63.117(a)(4)(i) and the text of S63.117(a)(4)(iv) were not
                              2-32

-------
consistent in presenting the monitoring requirements for
boilers and process heaters.Another commenter {A-90-19:
IV-F-7.33) claimed that, monitoring requirements for boilers
did not exist.
     Response;  Only boilers or process heaters smaller than
44 MW and combusting a process vent stream that is not used as
or mixed with the primary fuel are required to conduct a
performance test.  To clarify this requirement, the language
of the regulation text and the table have been revised so that
they are consistent.  Monitoring of firebox temperature is
also required for boilers meeting these specifications.
     No monitoring or testing is required of boilers 44 MW or
greater, or of those boilers below 44 MW that introduce the
process vent stream as the primary fuel or that mix the vent
stream with the primary fuel and introduce it through the same
burner.  The EPA decided that monitoring of these units was
not necessary because their burning characteristics would
ensure a 98 percent reduction in the organic content of the
process vent stream.  Monitoring for all other boilers below
44 MW is described in §63.114.
     Comment:  Two commenters (A-90-19:  IV-D-70; IV-D-99)
indicated that experience with catalytic incineration has been
problematic due to the degradation of the catalyst bed
(poisoning) that occurs under normal operation of this type of
control device.  As a result, the commenters (A-90-19:
IV-D-70; IV-D-99) recommended repeated performance testing to
detect catalyst poisoning and to verify the percent reduction
achieved.  The commenters (A-90-19:  IV-D-70; IV-D-99)
asserted that the performance testing should establish the
inlet catalyst bed temperature and the VOHAP and TOC
concentrations in the outlet stack gas, while another
commenter (A-90-19:  IV-D-34) claimed that the catalytic bed
inlet temperature is inconsequential to catalytic incinerator
performance and therefore need not be monitored.
     Response;  The temperature difference across the catalyst
bed has been determined to be sufficient for determining
proper operation of a catalytic incinerator, and additional
                              2-33

-------
performance tests would be unnecessary and burdensome.   A
change or drift in this temperature differential would
generally indicate cases of catalyst poisoning.   During the -
initial performance test, owners or operators of catalytic
incinerators must establish a site-specific parameter range
for temperature difference across the catalyst bed.  This
established range becomes their operating requirement.   The
owner or operator would be required to continuously monitor
inlet and outlet bed temperature and calculate the temperature
difference.  If the temperature difference is ever outside the
established range, this would be a violation of the operating
standard.
2.4.2  Monitoring
     Comment;  One commenter (A-90-19:  IV-D-92) requested
that alternative monitoring protocol, subject to the EPA's
approval, be allowed if the proposed monitoring requirements
are not feasible or economical for a particular facility.
     Response;  The EPA agrees and has provided for
alternative monitoring in §63.151(f).
     Comment;  One commenter (A-90-19:  IV-D-41) said there
should be clearly established emission limitations for process
vents and suggested that emission monitoring at the exit of
the last control device be performed regularly, no less than
annually.  The commenter (A-90-19:  IV-D-41) added that
product recovery devices or vapor collection devices should be
required for concentrated streams prior to any combustion
device.
     Response;  To allow for site-specific situations, the
regulation does not require mass emission limits for organic
HAP's (e.g., pound per hour limits).  Instead, the regulation
establishes a percent reduction limit.  The regulation lists
operating parameters to be monitored for each control device
and requires the source to establish site-specific parameter
ranges to ensure that the control device is properly
maintained and operated.  Continuous monitoring is required
for most controls.
                              2-34

-------
     If an owner or operator selects to monitor a parameter
that is not listed in the regulation, that owner-or operator
can request approval of such monitoring, -tO-include_ 	
establishing a range and monitoring frequency for the
parameter that would indicate proper operation of the control.
In most cases, parameter monitoring will be continuous rather
than annual.
     An owner or operator has the option of using a recovery
device to achieve a TRE greater than 1.0 or achieve a
98-percent reduction, but is not required to do so because a
recovery device may not meet the control TRE requirements in
all cases.
     Comment;  One commenter (A-90-19:  IV-D-73) supported the
monitoring exemption for boilers and process heaters that
introduce all vent streams with primary fuel, and one
commenter (A-90-19:  IV-D-73) strongly supported the
monitoring exemption for boilers and process heaters with
greater than 44 MW heat input.   To the contrary, two
commenters (A-90-19:  IV-D-85;  IV-F-7.33)  urged that
monitoring be required for all boilers and process heaters,
including those with heat input greater than 44 MW and those
introducing vents with or as primary fuel.  The commenter
(A-90-19:  IV-D-85) stressed that boiler or process heater
performance will decline if not operated and maintained
properly and said that the EPA should require monitoring of
these devices so that operators have the incentive to maintain
the devices properly and replace them before they deteriorate.
     Response;  The EPA agrees that there would be technical
and cost incentives to maintain the equipment properly because
boilers are usually used to generate heat and energy needed
for the process.  Sources must keep such boilers operating
properly in order to run their processes,  especially if the
vent stream is used as or introduced with the primary fuel.
Therefore, by reducing the monitoring requirements, the burden
on the facilities is also reduced.
     Comment?  one commenter (A-90-19:  IV-D-89) proposed
lowering the monitoring exemption for boilers and process
                             2-35

-------
heaters from less than 44 MW heat input to less than 8 MW heat
input.
     Response;  The 44 MW-was selected^based,on information
available during development of the SOCMI NSPS'On the  --  •--
temperature and residence times required by the boiler to
achieve the desired combustion efficiency.  The designs of
boilers larger than 44 MW are such that they would
consistently achieve over 98 percent reduction and monitoring
is unnecessary.  The commenter included no data to support
lowering the monitoring exemption for boilers to a heat input
capacity of 8 MW.  If the owner or operator does not want to
monitor temperature for the boiler, the owner or operator can
apply to monitor an alternative parameter.
     Comment;  One commenter (A-90-19:  IV-D-117) requested
that CEM's be used to measure THC at the inlet and outlet of a
thermal incinerator and that the monitoring data be submitted
on a monthly basis.  The commenter (A-90-19:  IV-D-117) also
suggested that for catalytic incinerators the THC and flow
rate be monitored continuously and the data be submitted
regularly.  Additionally, the commenter (A-90-19:  IV-D-117)
suggested that the flow rate to a flare be monitored
continuously and submitted monthly.  One commenter (A-90-19:
IV-F-10) requested better monitoring of release points,
including flares, to ensure they are working efficiently.
     Response;  HON requires semiannual reporting for most
plants, but quarterly reporting for those that are poor
performers.  This is frequent enough to enforce the standard
in a timely fashion, but it is less burdensome for sources  and
enforcement agencies than monthly reports  (see recordkeeping
and reporting BID volume 2E).  In addition, monthly reports
would increase the reporting and recordkeeping burden without
necessarily increasing any emission reduction benefits.  The
EPA considers temperature monitoring less burdensome than THC
monitoring, and adequate for compliance demonstration.
     Cp^m^TTfr1  One commenter (A-90-19:  IV-D-12) stated that
the wording in §63.114(a)(2) could be construed as excluding
infrared monitoring as a flare pilot flame detection device.
                             2-36

-------
The commenter's  (A-90-19:  IV-D-12) reasoning was that
thermocouple and ultraviolet monitoring were specifically
given as examples, which could be confusing.to_any persons
considering the application of infrared monitoring*  The
commenter  (A-90-19:  IV-D-12) suggested that either infrared
monitoring also be included as an example, or no specific
examples be given.
     One commenter (A-90-19:  IV-D-12) disagreed with the use
of the phrase "at the pilot light" in §63,114(a)(2) and
suggested that the phrase be removed from the above stated
paragraph or be rewritten to indicate that the thermocouple
needs to be "at the pilot light", but other sensors could be
remote while monitoring the pilot flame.
     Response;  The EPA did not intend to exclude infrared
monitoring as a possible flare pilot flame detection device.
In the regulation, thermocouple and ultraviolet monitoring
were mentioned as examples of possible detection devices;
however, other types of devices could be used as long as their
function is to ensure that the pilot flame remains lit.  For
this reason, infrared monitoring has been added into the
regulation as an example of a flare pilot flame detection
device.
     In addition, language has been added into the regulation
stating that the detection device used must ensure the pilot
flame is lit.  Such language would ensure that all possible
pilot flame detection devices, including those not on the
list, would achieve their desired function.
     Section 63.11(b)(5) of the General Provisions does not
include the phrase "at the pilot light."  If a thermocouple is
used as a detection device, then it must be positioned at the
pilot light to sense the flame.  However, there are other
detection devices, such as ultraviolet and infrared monitors,
that could be positioned remote from the pilot flame while
still monitoring its presence.
     To ensure that the wording of the regulation does not
preclude the use of infrared devices and other devices that
indicate the continuous presence of the flame but are not
                             2-37

-------
positioned at the pilot light to sense the flame,  §114(a)(2)
has been reworded as follows;
     (2) Where a flare is used> the following monitoring	
     equipment is:required:  a device (including;but not
     limited to a thermocouple, ultraviolet beam sensor,
     or infrared sensor)  capable of continuously
     detecting the presence of a pilot flame.
     Comment;  One commenter (A-90-19:  IV-D-54) pointed out
that scrubbers are not always used to control halogenated
streams and that in these cases, the scrubber is not
necessarily a recovery device.  If the scrubber controls a
non-halogenated stream, then pH is not an appropriate
monitoring parameter.  One commenter (A-90-19:  IV-D-70) also
suggested that the scrubbing fluid maintain a minimum pH
of 13.
     Response:  In the regulation, requirements are given for
three different control device scenarios that include
scrubbers.  These three different scenarios, their monitoring
requirements, and where the requirements are given in the rule
are outlined as follows:   First, for a halogenated stream
routed through a combustion device and then a scrubber,
monitoring of pH and L/G ratio is required  [see §63.114(a)].
Second, for a stream with a TRE between 1.0 and 4.0 that is
routed through a scrubber for recovery, liquid temperature and
exit specific gravity are monitored [see §63.114(b) for
absorbers].  And third, for a scrubber used in a configuration
other than the two mentioned above (including a non-recovery
scrubber used as a control device for a non-halogenated stream
and a scrubber used as a control device for a halogenated
stream prior to combustion), the provisions for monitoring
alternate control technologies must be followed.  In such
cases,  site-specific parameters must be selected and monitored
based on approval from the Administrator as described in
S63.ll4(c)(l) of the rule.
     Comment;  Two commenters  (A-90-19:  IV-D-70; IV-D-99)
suggested that performance testing and compliance parameters
for absorbers include solvent type and flow rate, specific
gravity of exiting solvent, and system pressure drop; but

                              2-38

-------
another commenter  (A-90-19:  IV-D-34) stated that monitoring
of scrubbing liquid specific gravity for absorbers^ was-
unnecessary.  Another commenter  (A-90-19:  IV-D-35) ^argued
that pH monitoring of the final scrubber effluent" is
sufficient for determining adequate scrubber performance.
This commenter (A-90-19:  IV-D-35) added that it  is
unrealistic to expect flow meters to withstand the harsh
conditions of both hot and acid service.
     Two commenters (A-90-19:  IV-D-79; IV-D-86)  stated that
gas flow to a scrubber is not necessarily a parameter that
indicates proper scrubber operation.  One commenter  (A-90-19:
IV-D-86) stated that scrubber liquid flow, scrubber pressure
drop, and pH are typically monitored to assure proper
operation and suggested that monitoring of either gas flow or
pressure drop be allowed in §63.114(a).
     Response;  The first three commenters are referring to
scrubbers used as recovery devices.  According to table 4 and
§63.114(b) of subpart G of the regulation, these  sources must
monitor specific gravity and exit temperature of  the absorbing
liquid.  However, provisions for monitoring of alternate
parameters are included in §63.114(c)(3) of the regulation.
If, based on site-specific conditions, an owner or operator
believes that it would be more appropriate to monitor a
parameter other than the ones mentioned above, then the owner
or operator could request approval from the Administrator to
monitor a different parameter.
     The next two commenters are referring to scrubbers used
following a combustor to control HC1 and other halogens and
hydrogen halides in the stream.  In this case, the owner or
operator is required to monitor pH and scrubber liquid/gas
ratio [see §63.114(a)  and table 3].  These monitoring
requirements are included in the rule to ensure that halogens
and hydrogen halides are being removed from the combustor
outlet.  A variety of flow meters constructed from different
materials are available for use in a caustic scrubber.  The
final two commenters are correct in pointing out  that gas flow
alone is not necessarily an indication of proper  operation.
                             2-39

-------
However, the intent of the regulation is to monitor the
scrubber liquid/gas ratio rather than the gas flow.  As
previously mentioned, ^63.114(c> allows owners or operators .to
apply to monitor an alternative parameter on a site-specific
basis.
     Comment;   One commenter (A-90-19:  IV-D-77)  maintained
that monitoring of the scrubbing liquid temperature and
specific gravity is not appropriate for non-recirculating or
"once through" scrubbers.  The commenter (A-90-19:  IV-D-77)
suggested that only liquid flow be monitored for a non-
recirculating scrubber.  One commenter (A-90-22:   IV-D-13)
recommended that the EPA allow alternate monitoring methods in
some situations where the methods already listed in the
regulation will not work.  The commenter (A-90-22:  IV-D-13)
stated that for absorbers temperature and specific gravity
must be monitored; however, if the organic content is very
low, there will not be sufficient changes in these parameters
to make them good indicators of absorber performance.  The
commenter (A-90-22:  IV-D-13) contended that a minimum
scrubbing flow that will achieve 98-percent efficiency can be
determined,  and that monitoring scrubbing flow should be
acceptable.
     Response;  The EPA assumes that if an absorber is used in
a recovery system, then the absorber recycles (or has the
potential to recycle) a portion of its effluent and is not a
once-through scrubber.  Furthermore, the EPA assumes that
absorbers used to scrub halogens from an incinerator's
effluent is a once-through scrubber.  As such, there are two
sets of monitoring and testing requirements for the two
absorber types just described.  For absorbers used in recovery
systems, a scrubbing liquid temperature monitor and a specific
gravity monitor are required, both with continuous
recordkeeping.  For absorbers used after an incinerator  (a
once-through scrubber), a pH monitoring device and flow meter
to measure scrubber liquid influent and inlet gas  flow rates
are required,  both with continuous recordkeeping.
                              2-40

-------
     As stated  in  §63.114(c) for process vents and §63.127(c)
 for transfer operations, owners or operators may request
 approval to monitor parameters other than Jthose listed in
 §63.l27(a) or  (b).
     The commenter (A-90-22:  IV-D-13) did not specify what
 they consider a "very  low" organic content.  However, if the
 exit organic content is 20 ppmv or less  (higher for streams
 that are originally greater than 1,000 ppmv), the compliance
 requirement is  being met.  A change of organic content within
 that range is irrelevant.  If the source or the permit
 authority has concerns regarding any monitored parameters,
 alternate parameters can be requested as specified in the
 above paragraph of this response.
     Comment;   Two commenters (A-90-19:  IV-D-70; IV-D-99)
 suggested that  exit stack gas VOHAP concentration be monitored
 continuously on regenerable carbon adsorption systems and  that,
 nonregenerable  systems utilize backup canisters in series  with
 primary canisters.  The commenters (A-90-19:  IV-D-70;
 IV-D-99) further suggested that for nonregenerative systems,
 hourly monitoring  occur between the primary and secondary
 canisters with  replacement of the primary canister in like
 kind when the primary outlet TOG concentration exceeds
 20 ppmv.  Another  commenter (A-90-19:  IV-D-33) suggested  that
 monitoring provisions be added to §63.114(b) and §63.127(b)
 providing the same opportunity for off-site regeneration of
 carbon canisters used  in a carbon adsorption system.
     Response;   The canisters referred to by the commenters
 (A-90-19:  IV-D-70; IV-D-99) are generally used on very small
 vent streams, batch processes, or small malodorous streams
 rather than on  continuous process vent streams subject to  HON
 monitoring requirements.  However, if such canisters are used
 on continuous vent streams subject to HON monitoring
 requirements, then the owner or operator could request
 approval to monitor alternate parameters as described in
 S63.114.
     Comment;   One commenter (A-90-19:  IV-D-113) stated that
while temperature  is an appropriate monitoring parameter for
                             2-41

-------
all types of adsorbers, specific gravity is not.  The
commenter (A-90-19:  IV-D-113) suggested that the EPA either
limit monitoring parameters for carbon adsorbers to
temperature only, or offer a secondary parameter appropriate
to the given technology, such as flow.
     Response;  Temperature and specific gravity monitoring
are required for absorbers, not adsorbers.  As discussed in
S63.114(d) and shown in table 4 of the proposed rule, carbon
adsorbers are required to measure regeneration stream
(e.g., steam) mass flow during regeneration and the
temperature of the carbon bed after regeneration.
     Comment;  One commenter  (A-90-19:  IV-D-64) requested
clarification of why the rule requires all the monitoring
records in table 7, §63.130(a)(2)(v) for carbon adsorber
regeneration stream flow and carbon bed regeneration
temperature, instead of requiring only daily averages.
     Response;  Detailed records are needed to assure that the
parameters remain within their established range.  A daily
average is not used because flow and temperature after
regeneration pertain only to the regeneration cycle, not to
other periods of operation during the day.
     Comment;  One commenter  (A-90-19:  IV-D-35) requested
that the EPA clarify the provisions in §63.118(b) requiring
that readily accessible records be kept for a product recovery
device or other means to achieve and maintain a TRE index
value greater than 1.0 but less than 4.0.  The commenter
(A-90-19:  IV-D-35) asserted that the phrase "or other means"
is confusing and should be clarified by stating that if a
process has a TRE value greater than 1.0 without using
recovery devices (i.e. absorbers, condensers, etc.)  it is
exempt from these provisions, or the phrase "or other means"
should be deleted from the section.
     Response:  A facility with a process vent stream
achieving a TRE greater than  1.0 without using a recovery
device (e.g., due to inherent process design or a process
modification) would have to apply to monitor and report a
                              2-42

-------
site-specific parameter under §63.114(c) and is not exempt
from monitoring provisions.
     Comment;  Two. commenters (A-90-19:  IV-D-70; IV-D-99)
recommended that, due to process fluctuations, Group 2 vents
with TRE greater than 4.0 should be allowed to follow the
monitoring provisions for Group 2 vents with TRE values
between 1.0 and 4.0.  Another commenter (A-90-19:  IV-D-48)
opposed the monitoring requirements for Group 2 vents with TRE
between 1.0 and 4.0.  The commenter (A-90-19:  IV-D-48)
reasoned that the requirements are burdensome and unnecessary
because reevaluation is required to determine if group status
changes when a process change is made.
     Several commenters (A-90-19:  IV-F-1.1 and IV-F-3;
IV-D-32; IV-D-48; IV-D-58; IV-D-62; IV-D-63; IV-D-69; IV-D-83;
IV-D-92; IV-D-112; IV-D-113) suggested that monitoring,
recordkeeping, and reporting requirements for Group 2 process
vents be reduced or eliminated.  In particular, three
commenters (A-90-19:  IV-D-32; IV-D-48; IV-D-69) thought it
was overly burdensome for Group 2 process vents to be required
to perform the same continuous monitoring, recordkeeping, and
reporting as Group 1 vents, because:  the Group 2 emission
points have been judged not to require additional control; the
rule requires TRE to be re-evaluated when process changes are
made; and the 112(g) modification program,, residual risk
determination, and future reviews of MACT standards are
sufficient to require future control of Group 2 vents, if
warranted.  Two commenters (A-90-19:  IV-D-58; IV-D-62)
suggested that reporting for these Group 2 process vents needs
to be limited to notification of changes that may potentially
lead to Group 1 designation of the source and develop
reporting and testing requirements for sources altered or
changed into potential Group 1 sources.  Two commenters
(A-90-19:   IV-D-32; IV-D-113) also supported the requirements
to monitor Group 2 process vents only where a recovery device
is used to maintain Group 2 status (TRE between 1.0 and 4.0).
     Response;  The rule requires monitoring, recordkeeping,
and reporting requirements for Group 2 vent streams with TRE
                             2-43

-------
index values between 1.0 and 4.0 to ensure those vents do not
become Group 1 vents due to process or recovery device
operating variations and- remain uncontrolled.  Group 2 vents
with TRE index values greater than 4.0 are not requixed to be
monitored.  An analysis was performed prior to proposal that
shows that a vent with a TRE greater than 4.0 is unlikely to
become Group 1 due to process or recovery device operating
fluctuations or measurement uncertainties, whereas if a vent
has a TRE is less than 4.0, it is important to monitor
recovery device operating parameters,  because variations in
process or recovery device operations could cause such streams
to become Group l.  The EPA recognizes the uncertainty present
in TRE calculations, but decided that a large enough safety
factor has been included for vents with a TRE greater than
4.0.  The decision not to require monitoring for process vents
with a TRE greater than 4.0 reduces the burden of the
regulation for both the industry and regulatory agencies.
     Comment;  One commenter  (A-90-19:  IV-D-92) stated that
it is not reasonable to require extensive controls for Group 2
process vents because the classification of vents as Group 2
implies that they cause a minimal environmental concern.  The
commenter (A-90-19:  IV-D-92) then asked what is the value
added by extensive reporting requirements if there is no
problem.
     Response;  The EPA agrees that controls for Group 2
process vents are not reasonable, and accordingly, the
proposed regulation did not require controls for Group 2
vents.  Monitoring, reporting, and recordkeeping are required
for Group 2 vents to ensure a Group 2 vent does not become a
Group 1 vent and go unregulated.  A Group 2 vent with a TRE
greater than 4.0 is not required to monitor the vent, but must
still follow the specified reporting and recordkeeping
requirements.
     Comment;  One commenter  (A-90-19:  IV-D-34) opposed the
monitoring requirements for process vents that qualify as
Group 2 through the low flow or low concentration
determination and stressed that engineering  judgement should
                              2-44

-------
be adequate since facilities face penalties if the engineering
assessment is found through testing to be incorrect.
     One commenter  (A-90-19:  IV-D-78) proposed that if
engineering judgement were allowed for the purpose of Group 1
and Group 2 determination through either the low flow or low
concentration clause, a safety factor of 4 be used to exempt
streams from monitoring, i.e., for vent streams with flow
rates less than 0.00125 scmm or HAP concentration less
12.5 ppmv.
     Response;  The rule allows three means for determining
that a vent is Group 2:  TRE calculation, or measurement to
verify that the stream is below either the specified flow or
concentration level.  If an owner or operator wishes to use
engineering assessment to determine group status, they can
calculate the TRE based on estimates of flow, emissions, and
heating value of the vent stream.  If the TRE is greater than
4.0, such engineering assessment is sufficient.  If the TRE is
less than 4.0, tested measurements are required.  The low flow
and concentration levels are included to reduce the testing
burden for small vent streams.  They can test only flow or
concentration to show the stream is below the 0.005 scmm or
50 ppmv levels instead of doing all of the testing that would
be needed to determine TRE.  These choices provide sufficient
flexibility.  To further reduce the burden for Group 2 streams
that are unlikely to become Group 1, no monitoring is required
for streams with TRE's greater than 4.0 or streams that are
below the low flow or concentration levels.  Such safety
factors as the commenter refers to have already been included
in selecting the TRE cut-off value of 4.0 for testing and
monitoring.
2.5  WORDING OF THE PROVISIONS
     comment:  One commenter (A-90-19:  IV-D-68) recommended
that exclusions from regulatory requirements, such as
monitoring exclusions, be included in each relevant section
because the rule is an "enormous document.14  The commenter
included an example that the monitoring exemption for the low
                              2-45

-------
flowrate and concentration limits are presented in S63.ll3(f)
and  (g) but not in §63-114.
     Response;  The EPA believes that the monitoring
requirements are clearly laid out in the rule.  The EPA
believes that since there are no flowrate (not to be confused
with flow indicator) or concentration monitoring provisions,
there is no need to add the exemptions to the monitoring
section and increase the volume of the document with redundant
information.
     Comment ;  One commenter (A-90-19:  IV-D-33) asked that
the wording in §63.111 and §63. 114 (d) referring to "flow
indicators" be revised to "flow indication system," which
would include computerized flow-metering systems as well as
traditional "flow indicators".
     Response;  The intent of the regulation is not to
preclude the use of "flow indicator systems."  Rather, the
term "flow indicators" includes computerized flow metering
systems as well as traditional flow indicators.  The intent of
the term "flow indicators" is to determine whether or not a
flow is present, and record this on a continuous basis.  A
flow meter capable of measuring flow rate (e.g., scmm) could
be used but is not required, because the simple presence of
flow in a bypass line is sufficient to detect the bypass of a
control device.
     Comment ;  One commenter (A-90-19:  IV-D-33) reasoned that
the word "replacement" as used in §63. 115 (e) should be taken
to mean a replacement of equipment not in kind, as in the case
of changing equipment service or upgrading equipment, since
replacement in kind should not be considered a process change.
     Response ;  If the new equipment configuration is
identical to the original equipment configuration, and the
owner or operator can show by calculation that the parameters
are identical between the old and the new configurations, then
testing and recalculation of TRE would not be required.
               One commenter (A-90-19:  IV-D-33) indicated
that the term water vapor had been used incorrectly in a
description of Method 4 and should be revised.
                              2-46

-------
     Response;  The commenter is correct that the term was
used incorrectly.  The intent is to use Method 4 to measure
the moisture content of the stack gas.  The regulation has
been revised.
     Comment;  One commenter  (A-90-19:  IV-D-33) suggested
eliminating the word "or" from §63.115(d)(2)(v)(B).
     Response;  The regulation has been revised to eliminate
the word "or" from S63.115(d)(2)(V)(B) since it is
unnecessary.
     Comment;  One commenter  (A-90-19;  IV-D-33) agreed that
the definitions of closed-vent system, control device, process
unit, and process unit shutdown  should appear in both §63.111
and §63.161 but asked that the definitions be made consistent
or the same in both sections.
     Response;  The EPA agrees with the commenter (A-90-19:
IV-D-33) where the differences in the definitions are not
necessary.  The definitions of closed-vent system and process
unit shutdown has been made consistent in §63.111 and §63.161.
However, there are intentional differences in the definitions
of control device and process vent.  The difference in the
control device definition is necessary for the process vent
requirements and the difference  in the process unit definition
is based on the subpart in which the term is used.
     Comment;  One commenter  (A-90-19:  IV-D-73) stated that
production rate should be deleted as an example of a process
change in §63.115(e) because production rates are continuously
changing in most SOCMI processes.
     Response;  Section 63.115(e) specifies that if a process
change (including production rate) is within the range used to
determine inputs to the TRE calculation, the process change
does not have to be reported and the TRE does not have to be
recalculated.  In addition, unintentional temporary changes in
production rate are excluded from this process change
reporting requirement.  However, a change in production rate
outside the range used to calculate the TRE for the NCS could
result in a change in TRE.  For this reason, recalculation of
                              2-47

-------
the TRE would be required if a production rate change causes
the equation inputs to exceed their initial range.
     Comment;  One commenter (A-90-19:  IV-D-77)  requested
that the definition of "vent stream" in §63.Ill to be the same
as "process vent" in §63.101.
     Response;  The EPA agrees, and the definition of "vent
stream" in §63.111 was modified to refer to the "process vent"
definition in S63.101.
     Comment;  One commenter (A-90-19:  IV-D-77)  requested
that the requirements in §63.113 be reorganized to show that
proposed paragraphs 63.113(6), (f), and (g) have equal weight
and that a source may meet any one of the three criteria, not
all criteria.
     Response;  The EPA believes that the requirements for
§63.113 are clear as written in the proposal package; and they
remain the same for promulgation.
     Comment;  One commenter (A-90-19:  IV-G-4) recommended a
change to the definition of "reactor process."  The commenter
(A-90-19:  IV-G-4) expressed concern that the proposed
definition could be interpreted to include product treatment
in storage tanks.  The commenter (A-90-19:  IV-G-4) explained
that it is common practice to add hydrogen peroxide, sodium
borohydrate, or various inhibitors to storage tanks to
maintain product quality or stability.  The commenter
(A-90-19:  IV-G-4) claimed that, although a minor reaction may
take place between the product and the added material, the
emissions and control techniques are characteristic of storage
tanks.  The commenter (A-90-19:  IV-G-4) suggested that the
definition of reactor process should be clarified to exclude
the addition of materials to product storage tanks for quality
or stability.
     Response;  The intent of the reactor process definition
is to cover a unit operation in which one or more chemicals or
reactants are added and the molecular structures are altered
to form one or more new organic compounds.  Compounds formed
during a minor reaction with a treatment chemical described by
                              2-48

-------
the commenter (A-90-19:  iv-G-4) would be assumed to be
present as impurities.
                             2-49

-------
                      3.0  STORAGE VESSELS

3.1  EMISSION CONTROL TECHNOLOGY
     Comment:  One commenter (A-90-19:  IV-G-4) stated that
the EPA should include cooling as a compliance option for
storage vessels.  The commenter (A-90-19:  IV-G-4) explained
that this option would involve lowering the temperature of the
stored liquid so that the vapor pressure is below the Group l
applicability level.  The commenter (A-90-19:  IV-G-4)
reasoned that this control strategy would be similar to the
one allowing installation of a product recovery device
following a process vent to raise the TRE index.  The
commenter (A-90-19:  IV-G-4) stated that adding such a control
option would provide an important opportunity for pollution
prevention.
     Response:  The EPA agrees with the commenter (A-90-19:
IV-G-4) that lowering the storage temperature of a stored HAP
should be allowed to affect Group I/Group 2 determination.
Though the proposed rule did not specify cooling as a
compliance option, the proposed rule did allow cooling to be
used to lower the HAP's "maximum true vapor pressure" which,
in turn, would affect a storage vessel's Group I/Group 2
status.  As defined in the proposed rule, the "maximum true
vapor pressure" of a stored liquid is based on the storage
temperature of the liquid.  A facility that chooses to lower
the storage temperature of a liquid HAP in order to reduce the
HAP's maximum true vapor pressure below the Group 1
applicability level may treat the storage vessel as a Group 2
vessel and comply with the Group 2 requirements.
     CQ^F^fc:  Two commenters (A-90-19:  IV-D-32; IV-D-112)
recommended that the EPA modify the RCT requirements for
                              3-1

-------
control devices to specify 90-percent removal for storage
vessels at existing sources and 95-percent removal for storage
vessels at new sources.  Oner commenter (A-90-19:  IV-D-112)
claimed that most existing storage vessels with refrigerated
condenser units can only achieve 90- to 93-percent efficiency.
The commenter (A-90-19:  IV-D-112) claimed that the proposal
BID indicated that refrigerated condensers are not capable of
meeting the 95-percent efficiency requirement.  The commenter
(A-90-19:  IV-D-112) claimed that existing refrigerated
condensers would need to be replaced with new cascade two-
stage systems in order to achieve 95-percent efficiency.
Another commenter (A-90-19:  IV-D-32) made reference to a
company that claimed that existing condensers would need to be
replaced with new cascade two-stage systems in order to
achieve 95-percent efficiency, but the commenter (A-90-19:
IV-D-32) did not express concurrence or non-concurrence with
the claim.  The commenter  (A-90-19:  IV-D-32) rather stated
that many existing refrigerated condensers would need to be
replaced with new ones.
     One commenter  (A-90-19:  IV-D-32) stated that the EPA had
assumed and not demonstrated that refrigerated condensers that
were installed to comply with requirements for 80- to
93-percent removal efficiencies could achieve 95-percent
efficiency.  Both commenters  (A-90-19:  IV-D-32; IV-D-112)
claimed that the EPA did not consider or justify, as required
by section 112(d) of the Act, the cost impact to facilities
that would need to replace existing refrigerated condensers
with new condensers.  The commenters  (A-90-19:  IV-D-32;
IV-D-112) stated that this replacement cost is not warranted
based on the small increase in control efficiency.
     Response;  The impacts of the proposed HON regulation did
account for the cost to replace condensers currently achieving
80- and 85-percent control efficiencies with new condensers
that can achieve 95-percent efficiency.  Furthermore, in
assigning new 95-percent-efficient condensers to existing
model tanks that previously had 80- and 85-percent-efficient
condensers, the EPA assigned both one-stage and two-stage
                              3-2

-------
cascade condensers.  Multistage systems are discussed on
page 2-31 of volume IB of the proposal BID.  As described in
the proposal BID, the type of condenser that was assigned to a
storage vessel depended on the properties and concentration of
HAP's in the vent stream from the storage vessel.  Regarding
one commenter's  (A-90-19:  IV-D-112) assertion that the
proposal BID indicated that refrigerated condensers are not
capable of meeting the 95-percent efficiency requirement, the
EPA assumes that this is a misunderstanding on the part of the
commenter.  The EPA considers this assertion inaccurate.
     Regarding condensers that are currently achieving
90-percent control efficiency, in developing the proposed HON
regulation, the EPA had assumed that recovery devices required
by State regulations to achieve an emission reduction of
90 percent could actually achieve an emission reduction of
95 percent with only a small increase in operating cost, by
simply lowering the temperature of the coolant.  This original
assumption is documented in docket item A-90-19: II-B-6.
     After reevaluating the available information, the EPA has
concluded that not all condensers currently achieving 90-
percent control efficiency will be capable of achieving 95-
percent control efficiency.  While the EPA has determined that
many of these 90-percent efficient condensers could achieve
95-percent control efficiency simply by lowering the
temperature of the coolant, there will be certain instances
where adjusting the coolant temperature will not achieve the
required emission reduction, due to a characteristic of the
condensers or of the stored chemical (e.g., high vapor
pressure).
     The EPA recognizes, based on currently available
information, that requiring replacement of existing well-
operated and maintained control devices that meet the control
efficiency achieved by sources at the floor (i.e., 90 percent
emission reduction) would not be justified.  This additional
control was estimated to cost about $38,000 for each
additional Mg of emission reduction achieved by existing
sources.   Therefore, the EPA has provided an exemption in the
                              3-3

-------
final rule for control devices installed on a storage vessel
on or before December 31, 1992 achieving at least 90-percent
emission reduction.
     However, for those storage vessels with a control device
achieving less than the floor (i.e., less than 90-percent
emission reduction) or for fixed roof tanks not equipped with
any control device, the EPA maintains that it is more
economically efficient to require 95-percent control, which is
based on the existing requirements in 40 CFR part 60
subpart Kb, rather than 90-percent control.
     Comment;  One commenter (A-90-19:  IV-D-58) contended
that the proposed storage vessel provisions should allow
covers on access hatches and automatic gauge float wells to be
either attached or bolted when they are closed.  The commenter
(A-90-19:  IV-D-58) explained that some companies use attached
devices rather than bolts to attach a cover or lid to the
roof.
     Response;  The EPA will allow fastening devices in place
of bolts to fasten hatches if these devices provide complete*
compression of the gasket when in use.  The wording in
S63.119(b)(6) and S63.119(c)(2)(ii) of the storage provisions
has been changed, as follows,  to allow these fastening
devices:  "Covers on each access hatch and each gauge float
well shall be bolted or fastened so as to be air-tight when
they are closed."
     Comment;  One commenter (A-90-19:  IV-D-58)  asserted
that the proposed requirement that "each roof drain" be
provided with a slotted membrane fabric cover that covers at
least 90 percent of the area of the opening should apply only
to "emergency overflow roof drains".  The commenter  (A-90-19:
IV-D-58) reasoned that a standard roof drain under normal
operation does not contain product and should be left
unrestricted at all times.
     Response;  The EPA interpreted this comment (A-90-19:
IV-D-58) to mean that a roof drain that does not contain
product does not drain into the stored liquid but, rather,
drains to a location outside of the storage vessel.  The EPA
                              3-4

-------
agrees with the commenter (A-90-19:  IV-D-58) that a roof
drain that drains to the outside of the tank should not
require a slotted membrane.  The wording of §63.119(c)(2)(vi)
of the storage vessel provisions has been changed to reflect
this exemption.  The phrase "each roof drain" was changed to
read "each roof drain that empties into the stored liquid."
     Comment;  One commenter (A-90-19:  IV-D-58)  asserted that
the storage vessel provisions should provide specifications
for how far a metallic shoe seal on an EFR should extend into
the stored liquid.  The commenter  (A-90-19:  IV-D-58)  stated
that this specification was provided for how far the seal
should extend above the liquid and explained that the  lack of
specification about how far it should extend into the  liquid
could produce "misunderstanding of the concept in the  proposed
rule".
     Response;  The EPA determined that S63.120(b)(5)(i) of
the storage vessel provisions should not specify the distance
that the lower end of a metallic shoe seal on an EFR should
extend into the stored liquid.  The EPA did not provide this
specification in the proposed HON because the distance that
the metallic shoe seal extends into the stored liquid  has no
effect on emissions from the storage vessel.  Rather,  it is
only important that the lower end extends into the liquid or
it would be a vapor-mounted seal.  The distance a metallic
shoe seal must extend into the stored liquid varies according
to the seal mechanism.  A metallic shoe seal must extend into
the stored liquid far enough to allow the weighted mechanism
to be attached.
     The EPA did specify the vertical distance the upper end
of the metallic shoe seal must extend above the liquid surface
in order to ensure that adequate contact area is provided
between the shoe and the vessel wall.
     cojfliqepfr;  One commenter (A-90-19:  IV-D-58)  contended
that the proposed provisions specifying that there shall be no
holes, tears, or other openings in the shoe, seal fabric, or
seal envelope of the primary or secondary seals of an  EFR tank
                              3-5

-------
should be changed to allow fabricated holes in the seal for
the roof anti-rotation device, i.e., bazooka guide bar.
     Response; The EPA would like to clarify for the commenter
(A-90-19:  IV-D-58) that prefabricated holes are allowed to be
part of the seal for the purpose of installing the roof anti-
rotation device;  however, the storage provision in
§63.119(c)(i)(iii) that the seal be "continuous" requires that
any prefabricated hole be filled with the anti-rotation device
and that a gasket be installed around the anti-rotation device
at the prefabricated hole, in order to ensure that the seal is
"continuous".  As long as each prefabricated hole is filled
with and gasketed at the anti-rotation device, the
prefabricated hole is not considered to be a "hole" for the
purposes of the storage provisions  [e.g., paragraph (a)(4) of
§63.119 of subpart G], because the prefabricated hole is
sealed and would not result in significant HAP emissions.
     Comment;  Three commenters (A-90-19:  IV-D-60; IV-D-86;
IV-D-97) advocated the use of vapor balancing as a means of
eliminating working losses from storage vessels.  One
commenter (A-90-19:  IV-D-60) recommended that the EPA add a
new RCT for storage vessels combining vapor balancing with
incremental control of breathing losses to achieve a total
emission reduction of 95 percent.  The commenter (A-90-19:
IV-D-60) added that there should be no requirement to
demonstrate compliance of the vapor balancing system but that
the engineering demonstration in §63.120(d)(1)(i) of the
proposed rule be required for the control device used to
control breathing losses.  The commenter (A-90-19:  IV-D-60)
stated that including vapor balancing as an RCT would provide
industry with the flexibility to utilize cost-effective
technologies to achieve MACT.  The  commenter  (A-90-19:  IV-D-
60) also stated that this approach would allow industry to
utilize existing equipment and equipment with a design
criteria less than 95 percent.   The commenter  (A-90-19:  IV-
D-60) provided an example calculation to illustrate combining
vapor balancing with a separate control device for
incrementally controlling breathing losses.
                              3-6

-------
     Response;  The EPA will respond to two possible
interpretations of the commenters1  (A-90-19:  IV-D-60;
IV-D-86; IV-D-97) suggestions:   (1) vapor balancing should be
allowed within a tank farm; or  (2) vapor balancing should be
allowed during transfer between  a storage vessel and a
transportation vehicle.  Regarding the use of vapor balancing
within a tank farm, the EPA concluded that this is not a
practical way to achieve 95-percent reduction of total organic
HAP emissions.  Vapor balancing  within a tank farm would
require excessive monitoring and coordination to ensure that
each time liquid is pumped into  one tank, liquid is also
pumped out of another tank.  Therefore, the EPA did not allow
vapor balancing as an RCT for storage vessels in the HON.
     Regarding the use of vapor  balancing during transfer from
a storage vessel to a transport  vehicle, the EPA allows vapor
balancing in the transfer provisions.
     Comment:  One commenter (A-90-19:  IV-D-113) expressed
support for the storage vessel provisions allowing floating
roofs and capture devices as RCT. The commenter (A-90-19:
IV-D-113) considers these controls to meet pollution
prevention goals and to have little or no cross-media impact.
Another commenter (A-90-19:  IV-D-32) expressed support for
the EPA's proposal to allow facilities to comply with the
storage vessel provisions by installing IFR's or EFR's as an
alternative to installing a closed-vent system and control
device.
     Two commenters (A-90-19:  IV-D-86; IV-D-97) requested
that the EPA specifically designate pollution prevention
measures in the storage vessel provisions.  One commenter
(A-90-19:  IV-D-103) expressed concern that the proposed
storage vessel provisions are inconsistent with pollution
prevention goals.  The commenter (A-90-19:  IV-D-103) asserted
that flares and other treatment  systems to control collected
emissions should not be allowed  for new vessels and suggested
that closed-loop recovery systems be required for new storage
vessels.  The commenter (A-90-19:  IV-D-103) added that the
EPA should explore the possibility of requiring closed-loop
                              3-7

-------
vapor recovery systems for multiple new storage vessels since
the cost would be the same as installing floating roofs on the
individual new storage vessels.
     Two commenters (A-90-19:  IV-D-86; IV-D-97) suggested
that control of liquid level in a tank is another pollution
prevention measure that can eliminate working losses with
minimal investment and cost.  The commenters (A-90-19:
IV-D-86; IV-D-97) also suggested tying vents of tanks in a
farm together with overall inlet/outlet control and inert gas
blanketing.  The commenters (A-90-19:  IV-D-86; IV-D-97) added
that in the context of the proposal BID model,  using these
methods would reduce emissions by only 80 percent.  One
commenter  (A-90-19:  IV-D-97) explained that this was due to
other important operating practices not being recognized.  The
commenter  (A-90-19:  IV-D-97) implied that, if these other
operating practices were recognized, that vent gas blanketing
could achieve greater than 80-percent emission reduction.
     Response;  The EPA allows pollution prevention measures
in the HON storage provisions.  The storage provisions require
floating roofs as RCT's; floating roofs are also pollution
prevention measures.  The storage provisions also allow
condensers to be used to achieve 95-percent reduction of total
organic HAP emissions from storage vessels, and condensers are
closed-loop recovery systems.  The EPA considered the
possibility of allowing control of liquid level to achieve
emission reductions but concluded that control of liquid level
could not achieve 95-percent reduction of total organic HAP
emissions from storage vessels.
     Regarding the use of inert gas blanketing, the EPA did
not include this pollution prevention measure in the storage
provisions because the EPA has no data demonstrating that this
measure can reduce total organic HAP emissions from storage
vessels by 95 percent.  The commenters (A-90-19:  IV-D-86;
IV-D-97) did not provide any information or data to specify
that inert gas blanketing could actually achieve 95-percent
emissions reduction.  The EPA contacted the commenter for
                              3-8

-------
clarification on this issue, but the EPA did not receive any
clarifying information.
               One commenter (A-90-19:  IV-D-103) stated that
the HON provisions for existing storage vessels should include
specific requirements that slotted guidepoles contain a float
to reduce emissions.
     Response ;  The EPA agrees with the commenter 's  (A-90-19:
IV-D-103) statement.  The proposed HON storage provisions for
EFR vessels at new and existing sources do include the
requirement for slotted guidepoles to be equipped with a float
(see §63. 119 (c) (2) (viii) (B) of the proposed HON regulation).
This requirement has been retained in S63.119 (c) (2) (x) of the
final rule.
     comment •  one commenter  (A-90-19:  IV-D-103) contended
that the EPA is required by the Act to require submerged fill
pipes for new and existing storage vessels between 250 and
40,000 gallons that are not controlled via floating roofs or
closed-vent systems and control devices.  The commenter
(A-90-19:  IV-D-103) explained that Louisiana, which contains
16 percent of the country's SOCMI process units, has this
requirement for vapor pressures greater than 10.3 kPa, and
Texas, which has 34 percent of the country's SOCMI process
units, has the same requirements for storage vessels between
1,000 and 25,000 gallons and vapor pressures greater than
10.3 kPa.
     Response ;  Regarding storage vessels with capacities
under 10,000 gallons, the EPA did not collect data on control
levels achieved by these storage vessels and is not regulating
these smaller storage vessels.  Regarding vessels with
capacities between 10,000 and 40,000 gallons, these vessels
are not splash-filled; rather, submerged fill is the standard
practice and represents the baseline level of control.
Therefore, the EPA concludes that there would be no emission
reduction benefits from the suggested requirement and that the
additional requirement would represent only an additional
recordkeeping burden.
                              3-9

-------
               one commenter (A-90-19:  IV-D-103) asserted
that the HON must require that all new storage vessels have
both liquid-mounted primary seals and secondary seals, and
that all new vessels not using a closed-vent system and
control device have welded rather than bolted deck seams and
no column penetrations.  The commenter (A-90-19:  IV-D-103)
contended that these additional controls are required to meet
MACT as prescribed in the Act,  which is "the maximum degree of
reduction in emissions" achieved by the "best controlled
similar source."  The commenter (A-90-19:  IV-D-103) further
asserted that the EPA is obligated to investigate the
possibility that 12 percent or more of existing sources meet
the emission level achieved with these additional controls and
to require these more stringent controls for existing vessels
if at least 12 percent are currently achieving these higher
control levels.
     Response;  The EPA has concluded that these additional
controls are not required as the MACT floor for new or
existing sources and are not incrementally cost-effective
enough to require above the MACT floor.  There is no existing
standard that requires welded deck seams, no column
penetrations, liquid-mounted primary seals, and secondary
seals; the EPA has not located any one source that implements
all four additional controls.
3.2  IMPACTS ANALYSIS
3.2.1  Cost Impacts
     (-orient;  One commenter (A-90-19:  IV-D-41) stated that
the HON cost estimates did not account for the cost savings in
products or reactants associated with the application of
recovery devices or the application of more efficient controls
such as floating roofs on storage vessels.  The commenter
(A-90-19:  IV-D-41) explained that failure to account for  such
cost savings prejudices the cost analysis toward
overestimating costs.
     Response; The EPA agrees with the commenter (A-90-19:
lV-D-41) that the cost savings incurred by the use of recovery
devices and floating roofs should be included in the cost
                             3-10

-------
analysis of the HON.  The EPA did include this cost savings in
the cost analysis for applying recovery devices
(i.e., condensers.and floating roofs) on storage vessels.  As
described in the proposal BID (BID volume 1C, section 4.4,
pp. 4-17 and 4-20), the value of the recovered chemical was
either the actual market price of the chemical or, if the
market price was not available, a default value for the
average chemical price (i.e., $1.57/kg), and the value of the
chemical recovered was subtracted from the total cost of the
recovery device.
     Comment;  Three commenters (A-90-19:  IV-D-32; IV-D-86;
IV-D-97) urged the EPA to use more accurate investment
estimates when evaluating costs of control systems.  The
commenters (A-90-19:  IV-D-32; IV-D-86; IV-D-97) recommended
that the EPA redo its cost analysis because, in their opinion,
the EPA had underestimated capital investment for control
systems by a factor of 2 to 5.  Two commenters  (A-90-19:
IV-D-86; IV-D-97) stated that because the annual cost
effectiveness for storage vessels is sensitive to investment,
it is important that the EPA use more accurate investment
estimates.
     Three commenters (A-90-19:  IV-D-32; IV-D-68; IV-D-97)
provided the same data indicating that the EPA's estimate of
the capital costs for installing an IFR on an existing fixed
roof storage vessel is low for two reasons: (1) the EPA's
estimated capital cost,  which is based on vendor quotes, is
lower than the vendor quote obtained by the commenters, and
(2) in general, vendor quotes underestimate the installation
cost for IFR's because they do not account for additional tank
repairs (i.e., upgrading column supports) that are discovered
after a tank has been emptied for the retrofit.  The data
provided by two commenters (A-90-19:  IV-D-32; IV-D-97)
indicates that for the capital cost of purchasing and
installing a 100-foot diameter fiberglass IFR on a 2,000,000
gallon tank containing methanol, the EPA-estimated capital
investment is $50,000, whereas the vendor-quoted cost was
$150,000 and the actual cost was $250,000. The data provided
                             3-11

-------
by two cor anters (A-90-19:  IV-D-32; IV-D-68) indicate that
the cost , ;r installing an IFR on a 1,500,000 gallon storage
vessel containing methyl methacrylate is $180,000 (excluding
the cost of tank repairs performed to accommodate the IFR).
     One commenter (A-90-19:  IV-D-97) asserted that the EPA
underestimated the cost to install a control device on a fixed
roof storage vessel,  because the EPA did not account for
start-up costs such as "prove-out" and "haz-op".
     Response;  The EPA considered the comment that EPA's
capital cost estimates for installing an IFR into an existing
fixed roof storage vessel are lower than the commenters'
(A-90-19:  IV-D-32; IV-D-68; IV-D-97) vendor-quoted estimates.
The EPA determined that the difference in estimates for
capital costs is due to the type of IFR being costed.  The
EPA's cost estimate of $50,000 is for a 100-foot diameter
aluminum IFR, while the vendor-quoted estimate of $150,000
provided by the commenters (A-90-19:  IV-D-32; IV-D-97) is for
a 100-foot fiberglass IFR, which is much more expensive.  The
EPA speculates, due to lack of information provided by the
commenter (A-90-19:  IV-D-68), that the vendor-quoted estimate
of $180,0.00 provided by the commenter (A-90-19:  IV-D-68) is
for a fiberglass or steel IFR, rather than for an aluminum
IFR.  The EPA's cost estimates are based only on aluminum
IFR's because, as described in the cost analysis in the
proposal BID, IFR's were assigned to storage vessels only when
the stored chemical would be compatible with aluminum.
Therefore, the EPA will not change its vendor quotes for
aluminum IFR's.
     The EPA's cost analysis indirectly accounts for the added
expense of installing a fiberglass or steel IFR, in those
cases where an aluminum IFR would be incompatible with the
stored chemical.  In its cost analysis, the EPA assigned and
costed aluminum IFR's only for those existing storage vessels
that contained compounds that would nc . be corrosive to
aluminum, such as halogenated chemicals and some glycol
ethers.  For all vessels containing chemicals corrosive to
aluminum, the EPA assigned and costed refrigerated condensers,
                              3-12

-------
rather than fiberglass or steel IFR's. The cost of
refrigerated condensers is comparable to the cost of
fiberglass and steel IFR's.  For example, for_an. 5S-foo.t_
diameter vessel storing a chemical with a vapor pressure of
1.51 psia, the EPA's cost estimate for installing a condenser
is $83,500.  For a 32-foot diameter vessel storing a chemical
with a vapor pressure of 7.12 psia, the EPA's cost estimate
for installing a condenser is $264,000.  These costs for
condensers are comparable with the cost of installing
fiberglass or steel IFR's reported by the commenters ($130,000
to $180,000).
     Regarding the comment that the EPA should not use vendor
quotes because actual costs are higher than vendor quotes due
to additional repairs, the EPA determined that these
additional repairs are not necessarily typical for installing
IFR's.  The EPA's cost estimate for installing IFR's includes
the cost of those repairs typically necessary to convert a
fixed roof tank to an IFR tank (i.e., the cutting of vents and
openings for modifying a vessel).  The commenter (A-90-19:
IV-D-97) provided a list of additional repairs and their
associated costs, including, for example, upgrading the leg
supports of a floating roof for a cost of $100,000.  However,
the EPA concluded that the additional repairs suggested by the
commenter are not typically necessary for retrofitting a
storage vessel with an IFR.  In some cases, these repairs
would need to be performed regardless of the IFR retrofit.
Therefore, the EPA considers its vendor quote for installing
IFR's to be valid.
     Regarding the comment that the EPA underestimated the
cost of installing a control device on a fixed roof storage
vessel due to the added costs for equipment start-up, the EPA
agrees with the commenter (A-90-19:  IV-D-97) that the
installation cost for condensers, which was the only type of
control device included in the EPA's cost analysis for storage
vessels, should reflect additional costs for start-up.  At
proposal, the EPA had utilized the costs provided in chapter 8
of Supplement 1 to the EPA's OCCM, Fourth Edition, PB92-
                             3-13

-------
137181, November 1991 for installing packaged (i.e., non-
custom) refrigerated condensers.  In reviewing the OCCM's
costing equations for condensers,  the EPA determined that the
cost equation for non-packaged refrigerated condensers
includes more of the start-up costs (e.g., for testing the
equipment after installation) than the cost equation for
packaged systems.  The EPA concluded that these additional
start-up costs should be accounted for, because sources will
be required to test their new equipment to ensure that it
consistently operates as described in the design evaluation
provided to the implementing agency for compliance purposes.
Therefore, the EPA has changed the equation for the
installation cost to include start-up costs by utilizing the
equation provided in the OCCM for non-packaged condensers.
     Comment;  One commenter (A-90-19:  IV-D-97) contended
that the price of the product used in the model tank in the
proposal BID is not a reasonable value.  The commenter
(A-90-19:  IV-D-97) noted that the price for methanol, which
is a common chemical used throughout the industry, is one-
third the value used for the model tank product.
     Response;  The EPA agrees with the commenter that the
price of methanol is less than the price of the example
product indicated in the proposal BID.  As indicated on
pages 4-17 and 4-20 of section 4.0 of volume 1C of the
proposal BID, in its analysis, the EPA used chemical prices
specific to the chemicals stored in the model tank farms
whenever they were available.  In the case of methanol, EPA
used a price of $0.35/kg.  For those chemicals for which no
price was available, such as for the model tank farm described
referred to by the commenter (A-90-19:  IV-D-97), an average
price of $1.57/kg was used.
     Comment;  Two commenters (A-90-19:  IV-D-32; IV-D-97)
suggested that the cost :-. the proposal BID for cleaning a
tank for conversion is Ic,.  One commenter (A-90-19:  IV-D-97)
indicated that the cost seems to be based on conversion of
gasoline or light petroleum products and therefore
underestimates cleaning costs for other products.  The
                             3-14

-------
conunenter  (A-90-19:  IV-D-97) stated that cleaning a
50,000-gallon tank at a specific site costs  $100,000.  The
conunenter  (A-90-19:  IV-D-97) stated ^that this x:ost is mere
than the EPA's estimate.  One conunenter  (A-90-19:  IV-D-32)
compared the EPA's estimated cost of $13,000 for cleaning a
2,000,000 gallon vessel to two companies' actual cleaning
costs of $1,000,000  (cleaning and disposal)  for a
1,000,000 gallon vessel and $208,000 (cleaning and
repairs/changes to accommodate the new IFR)  for a
1,500,000 gallon vessel.
     Response;  The EPA agrees with the  conunenters that the
cost of sludge disposal should be added  to the cost of
cleaning and degassing a storage vessel.  The EPA has added
the cost of sludge disposal, based on the assumption that
disposal will cost $5 per gallon of sludge and that each
storage vessel will contain two inches of sludge for disposal.
These assumptions are documented in the  EPA  document entitled
"Internal Instruction Manual for BSD Regulation Development:
Storage Vessels," Office of Air Quality  Planning and
Standards, Research Triangle Park, NC, January 1993.  These
costing assumptions result in a total cleaning and degassing
cost (including sludge disposal) of about $3,500 (1989
dollars) for a 40,000-gallon tank assumed to have a diameter
of 19 feet, and $26,400 for a 1,000,000-gallon tank assumed to
have a diameter of 60 feet.
     The EPA's revised estimates for cleaning and degassing
storage vessels, which includes sludge disposal costs, are
lower than the costs provided by the commenters.  However, the
cost estimates for cleaning and degassing storage vessels
provided by the commenters (A-90-19:  IV-D-32; IV-D-97) were
not substantial enough to justify the EPA's  changing its own
cost estimates, which were reviewed by tank  service companies
that handle SOCHI storage vessels.  The  cost estimates
provided by the commenters were not substantial enough
because:  (1)  they were not detailed enough  to explain why the
costs were high for the specific storage vessels mentioned
(e.g,  the nature of the stored chemical); and (2) they only
                             3-15

-------
represented a couple of example tanks that do not necessarily
represent typical costs for the whole industry.  Therefore,
the EPA will continue, to utilize its own cost estimates for
cleaning and degassing storage vessels.
     Comment;  One commenter (A-90-19:  IV-D-97) stated that
the EPA failed to consider that tanks in the chemical
industry, unlike the gasoline refining industry, may have been
in a variety of chemical services since being built, and as a
result, when conversion is required, old nozzles, etc. may
have to be removed, upgraded, or replaced.
     Response; The EPA did not include the cost of removing or
upgrading nozzles on storage vessels in estimating capital
costs for installing IFR's because the EPA does not consider
upgrading nozzles to be a change that will be made for the
average fixed roof storage vessel retrofit.  Additionally, the
EPA does not anticipate that, for those vessels requiring a
nozzle upgrade, the cost will be significant relative to the
total capital cost of installing IFR's.
3.2.2  Emission Estimates
     Comment;  Two commenters (A-90-19:  IV-D-32; IV-D-62)
recommended that the EPA update the storage vessel equations
in the final HON to reflect the latest changes in the EPA
document entitled "Compilation of Air Pollutant Emission
Factors (AP-42)."  The commenters (A-90-19:  IV-D-32; IV-D-62)
noted that the EPA published a "Supplement E" to AP-42 in
October 1992 which contains a new Chapter 12 entitled "Storage
of Organic Liquids".  One commenter (A-90-19:  IV-D-62)
indicated that the new equations for fixed roof storage
vessels would impact the EPA's cost and emissions analysis for
storage vessels.
     Response; The EPA has determined that the 1992 version of
the AP-42 equations for fixed roof storage vessels would not
significantly impact the EPA's cost and emissions analysis for
storage vessels.  As discussed in a memorandum entitled,
"Review of API Publication 2518—Evaporative Loss from Fixed
Roof Tanks—Preliminary Technical Evaluation of New Emissions
Data and New Emission Factors," the EPA considered the
                             3-16

-------
difference between the 1992 and the 1985 AP-42 equations for
fixed roof storage vessels by calculating breathing loss
emissions with both sets of equations, using both actual test
parameters and default values.  After comparing the results of
the two sets of equations with the actual test results of the
breathing losses, the EPA concluded that the 1985 and 1992
AP-42 equations provide comparable predictions of breathing
loss for petrochemicals.
     Additionally, in comparing the 1985 and 1992 AP-42
equations, the EPA determined that the 1992 AP-42 equations
are more site-specific than the 1985 AP-42 equations and will
require sources to make more measurements of tank parameters
because the EPA cannot provide default values for some of the
variables in the 1992 equations.  Therefore, the EPA will
retain the 1985 AP-42 equations for storage vessel emissions
in the regulation.  However, the EPA has decided that sources
should be given a choice to use either set of equations for
the breathing losses from fixed roof storage vessels, by
incorporating by reference the American Petroleum Institute
Publication 2518:  Chapter 19, Section 1 - Evaporative Loss
from Fixed-Roof Tanks, Second Edition, October 1991.  Although
the EPA will allow sources to choose either set of equations
for breathing losses from fixed roof storage vessels, the
source must use the same set of equations for estimating both
credits and debits from storage vessels for emissions
averaging.
3.3  APPLICABILITY AND GROUP I/GROUP 2 DETERMINATION
3.3.1  Applicability
               One commenter (A-90-19:  IV-D-68) stated that
the vapor pressure threshold for Group 2 tanks should be set
above 1.0 psia.  The commenter (A-90-19:  IV-D-68) contended
that the EPA had underestimated the cost to retrofit a methyl
methacrylate tank with a floating roof. The commenter
(A-90-19:  IV-D-68) reported that such a retrofit would cost
$389,000 (including installation of a floating roof, tank
cleaning, and sludge disposal) and would result in emissions
reductions of 11.2 Mg per year, yielding a cost-effectiveness
                             3-17

-------
of $35,000/Mg.  The commenter (A-90-19:  IV-D-68) indicated
that this cost-effectiveness exceeds the $3,400/Mg listed in
the proposal preamble to justify the Group I/Group 2
applicability criteria,  The commenter (A-90-19:  IV-D-68)
concluded that in order to maintain the $3,400/Mg cost, the
vapor pressure cutoff should be set above 1.0 psia which is
approximately the vapor pressure of methyl methacrylate.
     Response;  The EPA wishes to clarify that cost-
effectiveness is based on annualizing capital costs over the
life of the equipment (i.e., 10 years).  The cost-
effectiveness value of $35,000 suggested by the commenter
(A-90-19:  IV-D-68) was not calculated in this manner.  Using
the values provided by the commenter (A-90-19:  IV-D-68), the
EPA calculated the cost-effectiveness for the single storage
vessel containing methyl methacrylate.   If the capital cost of
$389,000 is multiplied by 0.263 (0.163 for capital recovery;
0.04 for taxes, insurance, and administrative charges; 0.05
for maintenance charges; and 0.01 for inspection charges), the
total annual costs (without considering the cost savings
associated with the product saved) are $102,310.  The value of
the product saved  (assuming a product value of $1.28/kg) is
$l,280/Mg multiplied by 11.2 Mg of emissions reduced, which
yields $14,336/yr.  The annual cost savings from saved
product, $l4,336/yr,  is subtracted from $!02,3lO/yr, which
yields $87,974 as the net annual costs. The cost-effectiveness
is calculated by dividing the net annual costs, $87,974/yr, by
the estimated annual emissions reduction, 11.2 Mg/year, which
yields approximately $7,900/Mg.
     The EPA understands that the cost-effectiveness varies
for specific chemicals and specific storage vessels.  However,
the decision to establish a standard above the MACT floor is
based on the average cost-effectiveness, not on the cost-
effectiveness of an individual chemical.  The corrected value
of $7,900/Mg, based on the commenter's (A-90-19:  IV-D-68)
provided data is not unreasonably higher than the average
cost-effectiveness value of $3,400/Mg cited in the proposal
BID.  Therefore, the EPA will not revise the cost-
                              3-18

-------
effectiveness value .of ^$3^,400/Mg developed for the proposed
rule.  The EPA maintains that this is a reasonable cost for
controls above the MACT floor.
     Comment;  One conunenter  (A-90-21:  IV-D-1) requested that
the vapor pressure cutoff for determination of Group 1 status
of storage vessels at new sources be increased to 5.2 kPa for
vessels with capacity greater than 151 cubic meters.  The
commenter (A-90-21:  IV-D-1) questioned the need to require
control of vessels storing chemicals with vapor pressures
below 5.2 kPa, stating that emissions from such vessels would
be low.  The commenter (A-90-21:  IV-D-1) specifically
objected to the fact that formaldehyde storage would be
subject to the rule.  The commenter (A-90-21:  IV-D-1)
reasoned that closed-vent systems and control devices would be
very expensive, and floating roofs would be impossible to use
because formaldehyde is stored in heated and agitated vessels.
The commenter  (A-90-21:  IV-D-1) explained that the
concentration of HAP in the head space over a liquid with such
a low vapor pressure would be very low, offering no fuel value
such that thermal destruction would require the added expense
of supplemental fuel.  The commenter (A-90-21:  IV-D-1) also
stated that the use of wet scrubbers on low concentration vent
streams is generally not efficient, and consistent achievement
of 95-percent removal of total organic HAP emissions would be
unlikely.  The commenter (A-90-21:  IV-D-1) further reasoned
that, since floating roofs may not be used on formaldehyde,
those vessels located at a distance from control devices would
require investment in extensive piping and blower systems.
     Response;  In proposing the vapor pressure threshold of
0.7 kPa for large storage vessels at new sources, the EPA
considered the cost of control.  For storage vessels that will
store liquids incompatible with an aluminum floating deck
(e.g., formaldehyde), the EPA assigned and costed a closed-
vent system and control device (i.e., refrigerated condenser).
The EPA determined that the cost of requiring control devices
is reasonable, given the emission reduction achieved.
                             3-19

-------
     The commenter (A-90-21:  IV-D-21) did not supply any data
indicating that a control device would be more expensive for a
storage vessel storing formaldehyde-tban for a storage,vessel
storing other HAP liquids incompatible with aluminum.
Therefore, the EPA sees no need to change the proposed vapor
pressure cutoff of 0.7 kPa for large storage vessels at new
sources.
     Comment;  One commenter (A-90-19:  IV-D-97) asserted that
the vapor pressure threshold for large existing tanks should
be raised to 1.0 psia and for large new tanks to 0.5 psia.
The commenter (A-90-19:  IV-D-97) explained that the economic
analysis associated with the model tanks in the proposal BID
contain underestimates of cost and overestimates of the
savings.  Section 3.2.1 of this BID volume lists specific cost
examples provided by the commenter  (A-90-19:  IV-D-97).
     Response;  As discussed in section 3.2.1 of this BID
volume, the EPA revised some of the assumptions in the cost
analysis for storage vessels.  As a result, the average cost-
effectiveness of the selected vapor pressure threshold for
MACT for large storage vessels at existing sources increased
only slightly, from $l,500/Mg at proposal to $2,000/Mg at
promulgation.
     ccnrnnen-^;  One commenter (A-90-19:  IV-D-103) contended
that the EPA is required by the Act to regulate new and
existing storage vessels with capacities of 25,000 to
40,000 gallons containing organic HAP's with vapor pressures
of 10.3 kPa or greater, rather than 13.1 kPa or greater.  The
commenter (A-90-19:  IV-D-103)  explained that Texas, which
presumably has more than 12 percent of the nation's storage
vessels, requires control of such tanks.  The commenter added
that Congress did not intend for the EPA to consider "economic
efficiency" in regulatory decision-making.
     Response:  The EPA wishes to clarify that although Texas
has more than 12 percent of the storage vessels in the SOCMI,
it does not necessarily follow that these 12 percent of SOCMI
vessels all are in the size category between 25,000 and
40,000 gallons capacity.  The EPA found the MACT floor for
                             3-20

-------
both new and existing sources in the smaller size category to
be 13.1 kPa and determined any additional control above the
MACT floor not to be costreffective.  Contrary to the
commenter's statement, the EPA is directed to consider cost-
effectiveness when considering whether to establish a MACT
standard above the MACT floor.
3.3.2  Group l/Group 2 Determination
     Comment;  One commenter  (A-90-19:  IV-D-68) stated that
the EPA should include turnover rate as a criterion in the
determination of Group I/Group 2 status for storage vessels.
The commenter (A-90-19:  IV-D-68) claimed that the EPA
overestimated tank emissions and therefore underestimated the
cost per Mg of emission reductions by basing the analysis on a
worst case turnover factor.
     Response;  In order to include turnover rate in the
determination of Group I/Group 2 status for storage vessels, a
source would need to be able to ensure its turnover rate for
each vessel.  The EPA based the analysis on worst-case
turnover rates (e.g., ranging from 11 turnovers per year for
2,000,000-gallon tanks to 372 turnovers per year for
10,000-gallon tanks) because the EPA concluded that no
practical method exists for a source to ensure its turnover
rate.
     Comment;  Three commenters (A-90-19:  IV-D-79; IV-D-86;
IV-D-97) favored use of a mass emission limit and/or
concentration cutoff, as alternatives to storage vessel size
and vapor pressure, for Group I/Group 2 status determination
for the storage vessel provisions.
     Two commenters (A-90-19:  IV-D-86; IV-D-97) contended
that the cost of controlling smaller tanks with the specified
RCT is unjustified.  The two commenters (A-90-19:  IV-D-86;
IV-D-97) maintained that the HON proposal BID states that
emissions from smaller tanks are half the emissions from the
model tank.  Both commenters (A-90-19:  IV-D-86; IV-D-97)
considered concentration,  mass,  or flows more appropriate.
     Response;  The EPA had considered the option of using a
concentration threshold or mass emission limit as the format
                             3-21

-------
for the storage vessel standard;  however, -the EPA concluded
that this format either would result in higher cost or would
be technically infeasible.   -  -
     The EPA concluded that establishing an emission limit for
IFR storage vessels would be economically infeasible.
Equipping each storage vessel with a capture and stack system
would require that the vessel vents be sealed and that the
emissions be transported to a measurement system. In most
cases, the closure of the vessel vents would require the
vessel to be blanketed with inert gas to prevent the formation
of explosive flammable mixtures in the vessel or the
measurement system.  This would be economically impracticable
considering that the sole purpose of the system would be for
emissions testing.
     The EPA concluded that establishing an emission limit for
EFR vessels would be technically infeasible.  It is
technologically impossible to equip EFR vessels with a closed-
vent system because these vessels are open to the atmosphere.
Whereas it is possible to equip EFR vessels with fixed roofs,
such a change would convert the vessels into IFR vessels, and
the rationale for not establishing an emission limit for IFR
vessels would still hold.
     The EPA concluded that establishing an emission limit for
storage vessels that are controlled with closed-vent systems
and control devices would be impracticable due to the
considerable variability in mass emissions from fixed roof
vessels.  Mass emissions from these vessels vary as a function
of vessel capacity, vapor pressure of the stored liquid,
molecular weight of the stored liquid, and utilization rate of
the storage vessel.  Because of the wide variation in the
amount of emissions of HAP vapors, a mass emission limit could
not be selected that would be achievable on a worst-case basis
(i.e., large vessel capacity, high vapor pressure, and high
utilization rate), and at the same time would not allow the
construction of closed-vent systems and control devices that
are less effective than MACT for other vessels.
                              3-22

-------
     The EPA disagrees with the comment  (A-90-19:  IV-D-86;
IV-D-97) that the cost of controlling smaller tanks with the
specified RCT is unjustified.  In the.analysis to determine
the MACT floor level of control for storage vessels, the EPA
divided the population of model vessels  into three size
ranges:  small, medium, and large.  For  the small and medium
size ranges, the EPA concluded that both new and existing
storage vessels should be controlled at  the MACT floor level,
which is the least stringent level of control allowed by
section 112 of the Act.  The EPA chose not to establish
emission control requirements more stringent than the MACT
floor for new and existing small and medium storage vessels
because the costs were considered high given the very small
potential emission reductions.
     Regarding the comment (A-90-19:  IV-D-86; IV-D-97) that
the HON proposal BID states that emissions from smaller tanks
are half the emissions from the model tank, the EPA assumes
that this is a misunderstanding on the part of the commenter.
The EPA considers this statement inaccurate.
     Comment:  One commenter (A-90-19:   IV-D-58) objected to
the proposed provision that vapor pressure determination be
based on true vapor pressure.  The commenter (A-90-19:
IV-D-58) asserted that the vapor pressure value for
Group I/Group 2 determination should rather be based on a
normalized temperature value, such as 70 °F.  The commenter
(A-90-19:  IV-D-58) acknowledged that normalizing temperature
may require different vapor pressure values according to
geographic region to reflect temperature differences between
climates.  The commenter (A-90-19:  IV-D-58) reasoned that the
true vapor pressure could vary for a product depending on
storage conditions, thus pushing the maximum true vapor
pressure above a HON applicability threshold for vapor
pressure.  The commenter (A-90-19:  IV-D-58) reasoned that a
normalized value would avoid changes that temperature effects
could have on control requirements.
     Response;  The EPA concluded that one vapor pressure
value based on maximum true vapor pressure is the most
                          '   3-23

-------
practical format for the provisions.  The commenter's
(A-90-19:  IV-D-58) suggestion to establish regional vapor
pressure values according to a normalized temperature does not
address those storage vessels that store liquid HAP's below or
above the ambient temperature.  Furthermore, developing
multiple vapor pressure values according to geographic region
would add unnecessary complication to the storage provisions.
     The EPA considered the commenter's (A-90-19:  IV-D-58)
concern that, because temperature varies from year to year,
the mr imum true vapor pressure will vary enough to affect the
centre  requirements necessary to comply from year to year
with tne HON provisions.  However, the EPA has determined that
the local maximum monthly average temperature, by which the
maximum true vapor pressure is determined for the purpose of
Group I/Group 2 determination, does not vary enough year to
year to change the maximum true vapor pressure enough to
affect a storage vessel's Group I/Group 2 status or specific
control requirements under the HON.
               one commenter  (A-90-19:  IV-D-33) contended
that the proper basis for determining Group I/Group 2 status
is not the true vapor pressure of the organic HAP being stored
but rather the partial pressure of the component of concern in
the vapor above the liquid.  The commenter (A-90-19:  IV-D-33)
defined the partial pressure as equivalent to the vapor
pressure of the component of concern, multiplied by the
concentration of the component of concern in mole fraction,
multiplied by the activity coefficient for the component of
concern in the liquid mixture.  The commenter (A-90-19:
IV-D-33) suggested rewording for proposed S63.119 (a) (1) and
(2) and S63.120(b) (1) of subpart G.
     Response:  The EPA agrees with the commenter (A-90-19:
IV-D-33) that the regulation should be based on the
equilibrium partial pressure exerted by the total organic
HAP's in the stored liquid.  Both the proposed and promulgate;
versions of the HON include this specification in the
definition of "maximum true vapor pressure1* in S63.111 of
subpart G.
                              3-24

-------
3.4  COMPLIANCE
3.4.1  General
     Comment;  Regarding the compliance requirements for
closed-vent systems and control devices, one commenter
(A-90-19:  IV-D-89) contended that smaller combustion devices
with a minimum residence time of 0.5 seconds and a minimum
temperature of 560 °C should meet the 95-percent emission
reduction requirement.
     Response;  The EPA concluded that the minimum residence
time of 0.75 seconds and minimum temperature of 816 °C
specified in the storage provisions in §63.120(d)(1)(i)(B) for
enclosed combustion devices should be replaced with a minimum
residence time of 0.5 seconds and a minimum temperature of
760 °C.  The EPA is reducing the residence time and
temperature in the storage provisions for control devices in
order to make the storage provisions consistent with the
transfer and wastewater provisions in the HON.  In reviewing
the proposed provisions for the three emission points, the EPA
determined that two sets of minimum residence time and
temperature requirements had been proposed for the three
emission points, and both sets of requirements, were intended
to achieve greater than or equal to 95 percent emission
reduction.  The EPA chose the least stringent of the proposed
requirements for control devices utilized on storage vessels,
as specified above.
     This change to the minimum residence time incorporates
the commenter's ( A-90-19:  IV-D-89) suggested minimum
residence time.  However, the EPA concluded that it would not
reduce the minimum temperature to 560 °C.  The data currently
available to the EPA supports a minimum temperature of 760 °C,
but does not support or refute the commenter's suggestion of
560 °C.  The commenter (A-90-19:  IV-D-89) did not provide
supporting data to the EPA.
     Comment;  One commenter (A-90-19:  IV-D-78) requested
that the storage vessel provisions requiring a closed-vent
system and control device include the option of complying with
a 20 ppmv outlet concentration limit as an alternative to the
                             3-25

-------
95-percent reduction requirement.  The commenter (A-90-19:
IV-D-78) reasoned that the proposed provisions for other
emission points included an outlet concentration limit in
addition to a percent reduction requirement.
     Response;  The EPA did not specifically require a 20 ppmv
outlet concentration as an RCT for storage vessels because
this type of RCT would require a more rigorous analysis,
including a performance test.  The HON storage provisions
require only a design evaluation for any control devices
utilized to meet the 95-percent emission reduction standard.
The EPA maintains that there is no reason to increase the
stringency of the storage provisions by requiring a
performance test to comply with an RCT.  Furthermore, the EPA
would consider it unreasonable to demonstrate compliance with
a performance test because the flow of organic HAP into a
control device from a storage vessel is highly variable.
     The EPA interpreted the commenter (A-90-19:  IV-D-78) to
be concerned that a source would be required to submit two
separate initial compliance reports for a control device
utilized for both storage vessel and process vent emissions,
in which case the HON might appear to.require a source to
submit both a design evaluation for storage showing 95-percent
emission reduction and a performance test for process vents
showing 98-percent emission reduction.  The EPA does not
intend to require this redundant reporting.  The EPA has added
a provision to S63.120(d)(1) of the final rule that allows a
source to submit, in place of a design evaluation, the results
from a performance test that are submitted as part of the
Notification of Compliance Status for compliance with the
process vent, transfer, or wastewater provisions of the HON.
The results of the performance test must demonstrate that the
control device achieves the emission reduction level required
by the storage provisions.
     Comment;  One commenter (A-90-19:  IV-D-77) interpreted
the proposed provision in §63.119(b)(2) to preclude normal
operation of a storage tank's liquid level below that at which
an IFR rests upon its leg supports.  The commenter (A-90-19:
                             3-26

-------
IV-D-77) claimed that this requirement would amount to an
approximate 5 to 20 percent loss in working inventory capacity
in tanks ranging from 10 to 40 feet tall, assuming a typical
leg support height of 2 feet.  The commenter (A-90-19:
IV-D-77) indicated that the liquid level is not routinely kept
at the 0 to 2 foot level range and that operating at this low
level is not a wise use of equipment resources, but that a
small fraction of the operating time some tanks do fall to
this level by design because a storage vessel is an
inventorying vessel with the intent of matching variable flows
to steady flows.
     The commenter (A-90-19:  IV-D-77) contended that the EPA
had not fully considered the economic and environmental impact
associated with the provision.  The commenter  (A-90-19:
IV-D-77) contended that the EPA should re-evaluate these
impacts and include the cost of, and marginally increased air
emissions associated with, new tanks built to recover the lost
inventory capacity.  The commenter (A-90-19:  IV-D-77)
reasoned that the operating limitation in §63.119(b)(2) would
worsen emissions by its inclusion as opposed to its omission
from the rule and contended that the costs and operating
management burden associated with the provision were not
justified.
     Response;  The EPA has concluded that the provision in
the HON that requires that "the process of filling, emptying,
or refilling** a storage vessel shall be "continuous and shall
be accomplished as soon as possible" during times when "the
floating roof is resting on the leg supports" will not cause a
reduction in a source's inventory capacity and will not
preclude normal operation of a storage vessel.  As stated by
the commenter (A-90-19:  IV-D-77), normal operation of a
storage vessel does not involve routinely resting the floating
deck on the leg supports.  The EPA agrees with the commenter
(A-90-19:  IV-D-77) that routinely resting the deck on the leg
supports would not be a wise use of the equipment or storage
space.   The intent of the requirement is to ensure that all
facilities establish the standard operating practice of not
                             3-27

-------
resting the deck on the leg supports and minimize the amount
of time that the deck does rest on the leg supports.
     The EPA would like to clarify the wording of the
provisions, that the process of filling, emptying, or
refilling "shall be continuous and shall be accomplished as
soon as possible."  This phrase implies that the activity of
filling, refilling, or emptying a vessel must meet both
criteria (i.e., "continuous" and "as soon as possible").  The
phrase "as soon as possible" is included in the provisions to
account for situations where a source has difficulty with
continuously filling, refilling, or emptying a vessel.  The
interpretation of the phrase will ultimately be determined by
the implementing agency.  Additionally, the EPA added a note
prior to both paragraphs (b)(1) and (c)(3) of S63.119 of the
final rule that clarifies the meaning of this regulation
language, which states that the intent of the provisions is to
avoid having a vapor space between the floating roof and the
stored liquid for extended periods.
3.4.2  Routine Maintenance
     Comment:  Several commenters  (A-90-19:  IV-D-32; IV-D-33;
IV-D-34; IV-D-69; IV-D-73; IV-D-79; IV-D-86; IV-D-89; IV-D-92;
IV-D-97; IV-D-112; IV-D-113) (A-90-21:  IV-D-7; IV-D-17)
asserted that the allowance in the proposed storage vessel
provisions of 72 hours for routine maintenance of a closed-
vent system and control device is insufficient.  Several
commenters (A-90-19:  IV-D-34; IV-D-69; IV-D-79)  (A-90-21:
IV-D-7) made the general statement that repair of various
control devices, including flares, incinerators, boilers,
thermal oxidizers, and water scrubbers, would require more
than the 72 hours per repair or per year.  Three commenters
(A-90-19:  IV-D-34; IV-D-112)  (A-90-21:  IV-D-7) stated that
boilers and incinerators require time for a systematic
cooldown period, for obtaining safety approval to open and
enter the combustion chamber, and for start-up without
damaging the unit, in addition to the time required for
checking and replacing parts.  Three commenters  (A-90-19:
IV-D-32; IV-D-86; IV-D-97) stated that maintenance of a flare
                             3-28

-------
may require up to 7 to 10 days per year.  One commenter
 (A-90-21:  IV-D-17) indicated that rebricking a thermal
oxidizer occurs approximately once every three years and
requires at least one week.  One commenter  (A-90-21:  IV-D-17)
indicated that many States require an annual inspection of
boilers, and this inspection itself requires more than
72 hours.  One commenter  (A-90-19:  IV-D-89) stated that five
days or more would be required to empty a large storage vessel
for the purposes of routine maintenance due to limited
tankage.  One commenter  (A-90-19:  IV-D-113) indicated that,
for certain chemicals such as styrene, there are special
maintenance procedures that increase the time required to
perform the maintenance.  Three commenters  (A-90-19:  IV-D-32;
IV-D-34) (A-90-21:  IV-D-17) stated that 72 hours would be
adequate for many typical procedures; for replacing a burned
tip on a flare; for performing one maintenance incident for an
incinerator; or for maintenance of simple equipment such as
activated carbon beds.  One commenter (A-90-19:  IV-D-32)
stated that the HON rule should provide an incentive for
facilities to complete such maintenance procedures as quickly
as practicable.
     Several commenters  (A-90-19:  IV-D-32; IV-D-33; IV-D-34;
IV-D-69; IV-D-73; IV-D-79; IV-D-86; IV-D-89; IV-D-92; IV-D-97;
IV-D-112; IV-D-113) (A-90-21: IV-D-7; IV-D-17) suggested
various ways to change the provisions to accommodate the need
for additional maintenance time.  Several commenters (A-90-19:
IV-D-69; IV-D-86; IV-D-89; IV-D-97; IV-D-112; IV-D-113)
(A-90-21:  IV-D-7) suggested alternative limits for routine
maintenance, including 10 days, 1 week, and 148 hours, with 10
days being the most common suggestion.  One commenter
(A-90-19:  IV-D-34) noted that a limit of 10 days per year for
maintenance has been incorporated by States into incinerator
air permits.
     Several commenters  (A-90-19:  IV-D-73; IV-D-89; IV-D-112)
(A-90-21:  IV-D-17) suggested allowing extensions for routine
maintenance.  Three of the commenters (A-90-19:  IV-D-73;
IV-D-112);  (A-90-21:  IV-D-17)  recommended allowing the State
                             3-29

-------
or local permitting authority to grant extensions.  One
commenter (A-90-19:  IV-D-89) stated that the HON should
include provisions for extensions.  Two commenters (A-90-19:
IV-D-89) (A-90-21:  IV-D-17) reasoned that allowing extensions
would be consistent with other air regulations, such as NSPS,
and current State agency practices which have proved effective
at granting variances within a reasonable time frame.
     Three commenters (A-90-19:  IV-D-32; IV-D-97) (A-90-21:
IV-D-17) suggested requiring that storage vessels not be
filled during periods of routine maintenance that exceed the
72-hour limit.  However, the commenters  (A-90-19:  IV-D-32;
IV-D-97); (A-90-21:  IV-D-17) stressed that sources should be
allowed to maintain a constant liquid level in the tank or to
lower the liquid level during this period.
     One commenter (A-90-19:  IV-D-33) suggested that the 72-
hour limit be applicable only to times during routine
maintenance when a storage vessel is being filled with organic
HAP.  The commenter (A-90-19:  IV-D-33) provided an example
calculation of breathing versus working losses indicating that
if working losses were restricted, total emissions based on
breathing losses would be very low.  The commenter (A-90-19:
IV-D-33) also indicated that if the EPA changed the 72-hour
limit to apply only to periods during routine maintenance when
a vessel is being filled with organic HAP, this 72-hour
allowance should be applied to and tracked for each storage
vessel separately in order to accommodate tank farms served by
one control device.
     Several commenters (A-90-19:  IV-D-86; IV-D-92; IV-D-97)
suggested that the 72-hour limit be replaced with a provision
that specifies performing maintenance as required by the
manufacturer of the control device.  Two commenters  (A-90-19:
IV-0-73; IV-D-79)  suggested that there be no limit for
rou- le maintenance.  One commenter (A-90-19:  IV-D-34)
sugc, .ited replacing the 72-hour limit with a specific limit
for each type of control technology, such as 240 hours per
year for incinerators.
                              3-30

-------
     Several commenters  (A-90-19:  IV-D-79; IV-D-86; IV-D-92;
IV-D-97)  (A-90-21:  IV-D-7) interpreted the 72-hour limit to
apply to  all maintenance,  including non-routine maintenance.
Two commenters  (A-90-19:   IV-D-86; IV-D-97) suggested that the
rule require documentation of all periods when equipment is
not operating properly, rather than establishing a limit.  One
commenter (A-90-19:  IV-D-79) suggested that the rule should
include provisions for unforeseen circumstances, such as
natural disasters, which would require control devices to be
out of commission for repair for periods of time in excess of
72 hours.  Three commenters (A-90-19:  IV-D-86; IV-D-92;
IV-D-97)  were concerned that the 72-hour limit would require
equipment to operate more than 99 percent of the time.
     Response;  The EPA wishes to clarify that "routine
maintenance," as it is referred to in the storage provisions,
refers to planned, routine maintenance of control devices,
excluding unplanned repairs due to malfunction.  Several
commenters (A-90-19:  IV-D-79; IV-D-86; IV-D-92; IV-D-97)
(A-90-21:  IV-D-7) interpreted the proposed 72-hour limit for
routine maintenance to apply to non-routine maintenance such
as malfunctions and repair due to natural disasters as well as
to routine maintenance.  The HON storage provisions include
provisions in §63.119(e)(5) of subpart G for control system
malfunction.   These provisions specify that in the event of a
closed-vent system or control device malfunction, the system
is not required to meet the specifications in §63.119(e)(1) or
(e)(2).  This provision implies that control systems are not
required  to operate more than 99 percent of the time, as three
commenters (A-90-19:  IV-D-86; IV-D-92; IV-D-97) stated.  The
EPA has clarified the language in the rule by changing each
reference to "routine maintenance" to "planned routine
maintenance".
     Regarding the issue of routine maintenance, the EPA has
concluded that the proposed 72-hour allowance for routine
maintenance is inadequate.  After reevaluating the available
information,  the EPA determined that increasing the time
allowance to 240 hours per year (i.e., 10 days per year) would
                             3-31

-------
be the most reasonable approach to address the need for more
time to complete routine maintenance, and would be consistent
with State air permitting activities.  The EPA did not choose
either of the other two approaches suggested by the commenters
because of the additional burden associated with them.
Specifically, requiring that storage vessels not be filled
during any routine maintenance exceeding 72 hours would
require the addition of equipment to monitor the liquid level
for enforcement purposes.  Further, allowing for extensions
for routine maintenance beyond the 72 hours would require
added reporting burden for both sources and implementing
agencies.
     While the EPA is allowing sources to utilize the full
240 hours to perform routine maintenance on each control
device, the EPA dos not expect that sources will utilize all
240 hours for all control devices, because many types of
control devices do not require this much maintenance time per
year.  The EPA has included provisions in the regulation that
make sources accountable for their utilization of this
allowance.  Sources are required to periodically (i.e., every
six months) report the routine maintenance performed during
the previous six months, including the amount of time used to
complete that routine maintenance, and the routine maintenance
that is planned for the following six months.
     Regarding the comment that it takes five days to empty a
tank for routine maintenance, for any storage vessel that must
be emptied before routine maintenance is performed
(i.e., storage vessels for which there are no organic HAP
vapor bypass capabilities), the owner or operator will not
need any allowance for time to perform routine maintenance
because there will be no additional emissions during the
routine maintenance operation.  The control device may be
operating while the tank is being emptied and degassed, and
once the storage vessel is emptied and degassed there will be
no emissions from the storage vessel.  Therefore, the length
of time of the allowance for routine maintenance does not
                              3-32

-------
apply to those cases where a storage vessel is emptied and
degassed for routine- maintenance of the control device.
     Comment :  One commenter (A-90-19:  IV-D-97) contended
that the storage vessel provisions should clearly state that
routine defrosting of refrigeration systems is not considered
part of the 72 hours of routine maintenance.  The commenter
(A-90-19:  IV-D-97) explained that this routine operation is
typically done during the morning hours for one hour, on a
daily basis, and is not done during the maximum representative
conditions.
     Response;  The EPA considers routine defrosting of
refrigeration systems to be a process integral to the
operation of a condenser unit rather than routine maintenance
for a condenser unit.  For those sources that choose to have a
backup unit to operate while defrosting the main condenser
unit, the defrosting cycle would not result in increased
emissions.  However, for those sources that do not choose to
have a backup condenser unit, the refrigerated condenser unit
must be designed to achieve a 95-percent reduction in total
organic HAP emissions, including the time that the unit is
undergoing the routine defrosting.  A source is required to
demonstrate any refrigeration system's 95-percent removal
efficiency in either a design evaluation or a performance
test.
3.4.3  Compliance Schedule
               Three commenters (A-90-19:  IV-D-32; IV-D-73;
IV-D-97) expressed support for the proposed provision to allow
an extension for upgrading seals for EFR's not meeting the
required standard as specified in 563 . 119 (c) (1) .  One
commenter (A-90-19:  IV-D-97) stated that because many process
units are served by only a few tanks, any tank outage could
shut down the process unit.  The commenter  (A-90-19:  IV-D-97)
indicated that allowing an extension for upgrading the seals
could alleviate scheduling problems.  One commenter  (A-90-19:
IV-D-97) stated that the environment would not be compromised
by this delay.  Another commenter (A-90-19:  IV-D-32)
explained that the emissions from emptying, cleaning, and
                             3-33

-------
degassing an EFR tank could exceed the emissions reduction
achieved by the seal upgrade.
     Four commenters (A-90-19:  IV-D-32; IV-D-58; IV-D-64;
IV-D-73) requested that the same extension allowed for EFR's
also be allowed for IFR's that have a vapor mounted primary
seal to retrofit a secondary seal or a liquid mounted primary
seal.  Two commenters (A-90-19:  IV-D-32; IV-D-73) explained
that a vapor mounted primary seal has been installed on many
vessels to comply with NSPS in 40 CFR subparts K and Ka and
State and local rules.   The commenters (A-90-19:  IV-D-32;
IV-D-73) asserted that the incremental benefits and costs of
making such an upgrade within 3 years of promulgation do not
justify this requirement.  One commenter (A-90-19:  IV-D-73)
maintained that if the EPA requires the seal upgrades within
only 3 years, the number of storage vessels involved, the
cost, and the disruption to plant operations and possibly
delivery will be greater than estimated in the proposal BID.
The other commenter (A-90-19:  IV-D-32) maintained that,
because the potential emission reduction from upgrading IFR
seals is lower than the potential emission reduction from
upgrading EFR seals (roughly IS percent compared to 25 to
30 percent), it should be reasonable to allow a similar
extension for both IFR vessels and EFR vessels.
     Two commenters (A-90-19:  IV-D-86; IV-D-97) stated that
the same extension should be allowed for conversion of any
tank to an IFR or EFR.
     One commenter (A-90-19:  IV-D-103) objected to the
proposed extension of up to 10 years for upgrading seals on
EFR's.  The commenter (A-90-19:  IV-D-103) explained that EFR
vessels have higher emission rates compared to IFR vessels and
that it is important that the seals on EFR's be upgraded as
expeditiously as possible.-  The commenter (A-90-19:  IV-D-103)
suggested allowing a maximum extension of five years after
promulgation for upgrading seals on EFR's.
     Response; Storage vessels are routinely emptied and
degassed on a 10-year cycle in order for the owner or operator
to conduct inspections for corrosion, weld failures, and
                              3-34

-------
standard API operating practices; and to remove sludge.  This
10-year cleaning and degassing schedule for performing storage
vessel inspections is part of the floor level of control for
storage vessels at existing sources.  If the final rule were
to require sources to upgrade the seal or fittings on an
existing IFR with an existing vapor-mounted seal within the
3-year compliance period, sources would not likely be able to
coordinate the activities of upgrading the seal or fittings
and performing the standard 10-year inspection.  As a result,
sources would be required to clean and degas a storage vessel
twice over a 10-year period, resulting in greater organic HAP
emissions than the emission reductions that would then be
achieved by the upgraded seals or fittings.
     As stated above, the logistics of inspecting and
servicing storage vessels according to the standard 10-year
schedule are too complex to accommodate changes to the 10-year
schedule.  The final rule will be affecting very large tank
farms, and each cleaning and degassing event will require that
tank farms be taken out of service over a period of time,
causing process unit shutdowns and affecting production
cycles.  Consequently., each additional cleaning and degassing
would represent a significant burden and added costs to the
SOCMI industry, which is already being required by the rule to
control large storage vessels at a level above the floor.
     For the reasons stated above, previous rulemakings have
allowed delays for upgrading the seals on floating roofs of
storage vessels at existing sources, with the intention of
avoiding premature storage vessel cleaning and degassing.  The
EPA wishes to uphold this intent in the final rule.  The EPA
has concluded that the 10-year extension provided in the
proposed rule for upgrading the seals of EFR's ia also
appropriate for IFR's that are equipped with a vapor-mounted
primary seal and no secondary seal [see $63.119(b)(3)(iv) of
subpart G of the final rule.]
     Regarding the comment (A-90-19:  IV-D-103) that the
extension for upgrading seals on EFR's in the proposed
§63.119(c)(1)(v)  of subpart G should be reduced from ten years
                             3-35

-------
to five years, the EPA would like to clarify that, as
explained above for IFR's, the lO'-year extension was chosen
because it is standard practice for sources to service storage
vessels every 10 years.  Requiring a shorter compliance period
such as the suggested five years would result in additional
HAP emissions from EFR storage vessels due to additional
storage vessel emptying and degassings.
     In the final rule, a 10-year extension is allowed only
for EFR storage vessels that are already equipped with one of
the following three seal configurations: (1) a vapor-mounted
primary seal and secondary seal; (2) a liquid-mounted primary
seal; or (3) a metallic shoe primary seal [see
§63.119(c)(1)(iv) and (c)(1)(v) of subpart G of the final
rule].   An extension is not allowed for EFR storage vessels
equipped only with a vapor-mounted primary seal and no
secondary seal.  In order to ensure that a single liquid-
mounted or metallic shoe primary seal and no secondary seal is
effectively reducing HAP emissions, the EPA is requiring in
the final rule that seal gap measurements of the primary seal
be performed at least once per year, rather than once every
five years.
     Comment:  Three commenters (A-90-19:  IV-D-58; IV-D-64;
IV-D-73)  stated that an extension should also be allowed for
the upgrading of IFR fittings, since controls cannot be
installed until the vessel has been emptied and degassed.  One
commenter (A-90-19:  IV-D-58)  specifically requested an
extension for equipping sample wells with slit fabric covers.
The commenter (A-90-19:  IV-D-58)  noted that, although
controlled fittings are currently required in 40 CFR part 60
subpart Kb, subpart Kb only applies to new tanks.  One
commenter (A-90-19:  IV-D-73)  stressed that it would be
environmentally and economically counterproductive to require
existing vessels equippe  vith IFR's in good condition to be
removed from service, c.  led, and degassed just to change a
deck fitting such as a •-  pie or gauging well or seal
mechanism within the 3-year compliance period.  The commenter
(A-90-19:  IV-D-73) argued that it would be more
                              3-36

-------
environmentally beneficial for available resources to be used
to install floating roofs where they do not currently exist
rather than to gain a small emission reduction by upgrading
seals and fittings.
     One commenter (A-90-19:  IV-D-73) requested that the EPA
revise the proposed requirement for providing projections
below the liquid level for openings on noncontact EFR's.  The
commenter (A-90-19:  IV-D-73) explained that the incremental
benefit of complying with this requirement for existing
storage vessels is not justified because the vessels must
first be removed from service and degassed.  The commenter
(A-90-19:  IV-D-73) suggested that the proposed provision
apply only to new or replacement EFR's or that an extension be
allowed for existing storage vessels until the next time the
vessel is out of service, or ten years after promulgation,
whichever is earlier.
     Response;  The EPA agrees with the commenters (A-90-19:
IV-D-58; IV-D-64; IV-D-73) that, if an extension is allowed
for upgrading seals on IFR's, then the same extension should
be allowed for upgrading the fittings for the IFR's, because
many of the fitting upgrades will require that the storage
vessels be emptied and degassed.  The EPA also agrees that an
extension should be allowed for providing projections below
the liquid level for openings on noncontact EFR's.  As
described above for extending the compliance schedule for
upgrading certain seal configurations on IFR's, the EPA
determined that it would be unreasonable to require a cleaning
and degassing within 3 years of promulgation solely for the
purpose of upgrading the fittings on an IFR or for providing
projections below the liquid level for non-contact EFR's,
because such requirements would result in additional organic
HAP emissions and additional costs.  The EPA added language to
the storage provisions in SS63.ll9(c)(5)(viii) and
63.119(c)(2)(xii) of subpart G to allow these extensions.
3.4.4  Inspections and Delay of Repair
     Comment;   One commenter (A-90-19:  IV-D-73) expressed
support for the delay-of-repair provisions in proposed
                             3-37

-------
§63.120(f)(2)(iii) of subpart G for storage vessel closed-vent
systems.  The commenter (A-90-19:  IV-D-73) explained that
most closed-vent ^systems, .cannot _be ^repaired .within -15 days
because most of them serve multiple emission points and
process equipment which must first be shut down.  One
commenter (A-90-19:  IV-D-78) implied that clarification was
needed regarding the actions that must be taken if a closed-
vent system has detectable emissions greater than 500 ppmv.
The commenter (A-90-19:  IV-D-78) suggested that the storage
vessel provisions appear to imply that if repair of such a
closed-vent system is not completed within 15 calendar days,
then the storage vessel must be emptied and degassed.  The
commenter (A-90-19:  IV-D-78) suggested that in cases where a
closed-vent system is found to have emissions greater than
500 ppmv, a delay of repair should be granted, upon reasonable
request, during which time a facility can continue storage and
use of tank contents, as long as no additional filling
activities occur.  The commenter (A-90-19:  IV-D-78) explained
that the greatest emissions occur during filling.  The
commenter (A-90-19:  IV-D-78) stated that if the EPA does not
allow such a delay and the storage vessel must be emptied, the
result would be an excessive ,-urden for facilities and
i. jssibly overall greater emir..: Lons.
     Response;  The EPA woulc iike to clarify that, in the
proposed provisions, delay of repair is allowed if one of the
following is true: (1) repair would be technically infeasible
without a process unit shutdown; or (2) emissions of purged
material resulting from immediate repair would be greater than
the fugitive emissions likely to result from delay of repair.
According to the proposed provisions, if one of those two
criteria were met, then repair could be delayed, upon request,
until the end of the next process unit shutdown.  If a source
does not meet one of these criteria, the source will not get a
delay for repair of the closed-vent system.  For these
sources, the EPA does not consider the cost of temporary
alternate storage capacity to be unreasonable if repair of the
closed-vent system cannot be completed within 15 days.
                             3-38

-------
     The delay of repair provisions have been moved to a new
section  (S63.148) of subpart G of the rule, which includes all
provisions concerning closed-vent systems that were previously
located in the storage provisions.
     Comment;  Two commenters (A-90-19:  IV-D-32; IV-D-113)
expressed support for the proposed provisions which would
allow two 30-day extensions to repair equipment failures that
are discovered during annual inspection of an IFR or 5-year
inspection of an EFR.  One commenter (A-90-19:  IV-D-32) gave
examples of circumstances that could prevent facilities from
being able to complete repairs during the initial 45 days:
(1) lack of alternate storage capacity and (2) safety
considerations such as tank shell corrosion restricting safe
access to the tank roof and exposure of repair personnel to
harmful vapors.
     Two commenters (A-90-19:  IV-D-92)  (A-90-21:  IV-D-1)
requested that the overall period for repair of floating roofs
be lengthened from 45 days plus two 30-day extensions to a
period of 90 days plus two 45-day extensions.  One commenter
(A-90-21:  IV-D-l) explained that the longer repair period
would ensure that facilities could do repairs during the
common down periods, Christmas and July 4th holidays,  one
commenter (A-90-19:  IV-D-92) explained that tank seals
typically cannot be obtained within 45 days from the suppliers
and there are storage/degradation problems if spares are
stored on site.  The commenter (A-90-19:  IV-D-92) added that
it would be very difficult to empty tanks because SOCMI
facilities do not have spare tankage.
     Response;  The EPA established the schedule of a maximum
of 90 days for completing repairs of floating roofs based on a
reasonable estimate of the time sources would require to order
and install needed parts.  This same schedule for repair is
included in the National Emission Standard for Benzene
Emissions from Benzene Storage Vessels (54 FR 38077,
September 14,  1989) in subpart Y of 40 CFR part 61 and the
Standards of Performance for Volatile Organic Liquid Storage
Vessels (52 FR 11429, April 8, 1987) in subpart Kb of 40 CFR
                             3-39

-------
part 60.  Based on discussions with vendors of IFR vessels,
parts for floating roofs are readily available and can be
ordered and received..within two-weeks» ~-The EPA-understands
that certain types of seals (i.e., liquid-mounted seals) are
bulky and require too much space for storage on-site.
However, since the lead time for ordering and receiving parts
is relatively short, it is not necessary for sources to store
these parts on-site.  For these reasons and because the EPA
received support for the proposed repair schedule, the EPA has
retained the proposed repair schedule in the promulgated
standard.
     Comment;  One commenter (A-90-21:  IV-D-1) urged the EPA
to reconsider the storage vessel provisions in proposed
§63.120(a)(7) requiring needed repairs discovered during
internal 10-year inspections of an IFR to be completed before
refilling of the storage vessel.  The commenter (A-90-21:
IV-D-1) contended that the storage vessel provisions should
allow refilling prior to repair with repairs to be completed
at a later date.  The commenter (A-90-21:  IV-D-1) explained
that, if spare parts were not readily available,  loss of
storage capacity could inhibit plant operation.  The commenter
(A-90-21:  IV-D-1) suggested that, if a repair could not be
made before refilling a vessel, the operator be required to
provide notice to the Administrator of the reasons the repair
could not be achieved prior to refilling and that the operator
then have six months to complete repair.  The commenter
(A-90-21:  IV-D-1) added that a six-month extension would
allow repairs to be conducted during common down times.
     Response:  As stated in a previous response in this
section, the EPA has concluded that spare parts for floating
roofs are readily available, and may be ordered and obtained
within two weeks.  If this short-term loss of storage capacity
would inhibit plant operation, a facility may choose to rent
additional tankage in the interim.  However, the EPA does not
anticipate this schedule to be a problem.  This same schedule
for repair is included in the National Emission Standard for
Benzene Emissions from Benzene Storage Vessels (54 FR 38077,
                                                  0
                              3-40

-------
September 14, 1989) in subpart Y of 40 CFR part 61 and the
Standards of Performance for Volatile Organic Liquid Storage
Vessels  (52 FR 11429, April 8, 1987) in subpart Kb of 40 CFR
part 60.  Additionally, the EPA would consider it unreasonable
for a source to empty and degas a storage vessel for its
10-year  internal inspection, to refill the vessel without
completing repairs, and then to empty and degas the vessel a
second time within six months to complete the repair.  This
additional degassing would be costly and would result in
additional HAP emissions.  For these reasons, the EPA is
retaining the repair schedule specified in proposed
§63.120(a)(7) for floating roofs.
     Comment;  One commenter (A-90-19:  IV-D-62) recommended
that internal inspections requiring emptying and degassing of
IFR vessels be required only once every 20 years, rather than
every 10 years.  The commenter (A-90-19:  IV-D-62) explained
that these internal inspections can be costly and dangerous,
and can  result in additional emissions of criteria pollutants
and HAP's from the degassed vessel.  The commenter (A-90-19:
IV-D-62) added that additional tankage may have to be built to
make up  for lost capacity.  The commenter (A-90-19:  IV-D-62)
also claimed that the cost to prepare a single storage vessel
can be as high as several hundred thousand dollars.
     The commenter (A-90-19:  IV-D-62) explained that a
20-year  inspection schedule is part of an industry-recommended
practice, included in American Petroleum Institute (API)
Bulletin 653, which requires that storage vessels undergo
internal inspections every 10 years, unless it can be
demonstrated that there are no corrosion problems, in which
case inspections take place every 20 years.  The commenter
(A-90-19:  IV-D-62) suggested that coordination of such
corrosivity inspections with the seal inspections of the HON
would provide an incentive for facilities to adopt improved
storage vessel management practices and would reduce the cost
of compliance.
     Response;  The EPA determined that storage vessels are
typically emptied, degassed, and cleaned every 5 to 10 years,
                             3-41

-------
and that the 10-year internal inspection requirement is not an
undue burden.  The EPA determined^ that, in many cases,
alternate 'tankage will be available.-  Otherwises-additional
tankage may be rented.  The EPA concluded that the suggested
20-year inspection schedule for vessels where the source has
demonstrated that there are no corrosivity problems would not
be an acceptable substitute for internal inspections, which
ensure that the floating deck and seals are operating
properly.
     Comment;  Two commenters (A-90-19:  IV-D-64; IV-D-73)
stated that the applicability for an exemption from inspecting
difficult-to-inspect closed-vent systems should be based on
the status of the entire closed-vent system, not on the status
of the storage vessel, as in proposed §63.120(f)(4)(ii).   One
commenter (A-90-19:  IV-D-73) explained that it is the closed-
vent system and not the storage vessel to which these proposed
provisions would apply.  One commenter (A-90-19:   IV-D-64)
referred to the exemption as a "temporary" exemption from
performing annual leak inspections.
     One commenter (A-90-19:  IV-D-73) suggested that this
exemption should apply to new closed-vent systems as well as
existing closed-vent systems.  The commenter (A-90-19:
IV-D-73) explained that certain components will continue to be
located in difficult-to-inspect locations on new closed-vent
systems because these design locations serve a purpose.  The
commenter (A-90-19:  IV-D-73) offered as an example that high
point bleeds, which are used for placing a piping system in
service or taking it out of service, are frequently difficult
to inspect, and this type of component, by definition, must be
located at the high point in the piping system, whether it is
a new or existing system.  The commenter (A-90-19:  IV-D-73)
further explained that the only way to make such a component
accessible, in some cases, is to build a very large platform
or scaffold.  The commenter  (A-90-19:  IV-D-73) contended that
there would be little environmental benefit from requiring
monitoring of these small components. The commenter  (A-90-19:
                              3-42

-------
IV-D-73) suggested that paragraph  (f)(4)(ii) of proposed
§63.120 be deleted from the storage vessel provisions.
     Response:  The EPA agrees with-the commenter  (A—90-19^
IV-D-73) that there are some types of equipment associated
with closed-vent systems that will be difficult to inspect
even in new sources.  The EPA will therefore apply the
difficult-to-inspect provisions for existing sources in the
proposed §63.120(f)(4) to storage vessels at new sources as
well.  The EPA has omitted the provisions that limited this
exemption only to storage vessels at existing sources.
Furthermore, in the final rule, the EPA has moved the
Method 21 leak inspection provisions for closed-vent systems
associated with storage vessels from the storage provisions
(i.e., §§63.119 through 63.123) to a new §63.148 of subpart G.
This new section includes the Method 21 leak inspection
provisions for closed-vent systems associated with the
transfer and wastewater provisions as well the storage vessel
provisions.  The provisions for difficult- and unsafe-to-
inspect parts are included in the new §63.148 provisions for
emission points at both new and existing sources.
     The EPA wishes to clarify for the commenter (A-90-19:
IV-D-64) that the exemption for difficult-to-inspect parts is
not a "temporary" exemption.  The exemption is associated with
a requirement to perform a less frequent inspection of the
closed-vent system.
     Comment;  One commenter (A-90-19:  IV-D-32) expressed
support for the provisions in §63.120(b)(7) of the proposed
HON allowing an extension of the inspection period when an
owner or operator determines that it is unsafe to perform the
seal gap measurements required under §63.120(b) for EFR
vessels.
     Response:  The EPA agrees and appreciates this support.
     Comment•  One vendor (A-90-19:  IV-D-8) provided
information to the EPA on a leak detection device to be used
instead of Method 21 for compliance with the inspection
provisions for closed-vent systems.  Others (A-90-19:
IV-D-14; IV-D-15;  IV-D-17; IV-D-18; IV-D-19; IV-D-20; IV-D-23;
                             3-43

-------
IV-D-24; IV-D-25; IV-D-27; IV-D-28), (A-90-20:  IV-D-2;
IV-D-4) commented on procedures and requirements of Method 21.
     Response;  The EPA has provided a discussion~on Method-21
and this alternate leak detection device in section 5.0 of
this BID volume.
     Comment;   Three commenters (A-90-19:  IV-D-32; IV-D-33;
IV-D-73) contended that the storage vessel provisions should
not include inspection requirements for closed-vent systems.
The commenters (A-90-19:  IV-D-32; IV-D-73) suggested that,
instead, only the requirements of subpart H should apply.  The
commenters (A-90-19:  IV-D-32; IV-D-73) added that, in order
to reduce confusion, the EPA should delete from the storage
provisions the references to closed-vent systems having no
detectable emissions.  One commenter (A-90-19:  IV-D-73)
suggested that the storage provisions state explicitly the
applicability of SS63.171 and 63.172 of subpart H for
monitoring all components of closed-vent systems for
detectable emissions.
     One commenter  (A-90-19:  IV-D-73) explained that in many
SOCMI plants,  there is one plant-wide closed-vent system and
that it is essential that the entire system be subject to only
one regulatory requirement.
     The commenter  (A-90-19:  IV-D-73) also noted a difference
between the inspection provisions for systems serving storage
vessels (subpart G) and those serving process equipment
(subpart H).  The commenter (A-90-19:  IV-D-73) stated that
§63.160(a) of the proposed subpart H sets a concentration
threshold of applicability for S63.172, whereas there is no
concentration threshold of applicability for inspecting
closed-vent systems under the proposed storage vessel
provisions.  The commenter (A-90-19:  IV-D-73) expressed
concern that for a single closed-vent system that serves both
storage vessels and process production equipment subject to
the equipment leak provisions, the closed-vent system that
does not meet the concentration threshold for the equipment
leak provision would be subject to inspection under the
storage vessel provisions.
                              3-44

-------
     One commenter  (A-90-19:  IV-D-32) questioned why closed-
vent systems for storage vessels were treated as different
from other types of equipment.subject.to.subpart_H_and  _ -
asserted that there was no reason why closed-vent systems on
storage vessels should have a separate requirement.
     Response;  The EPA reviewed the option of consolidating
all fugitive emission testing in subpart H. However, due to
the structure of subpart H and to the different compliance
schedules for subparts G and H, incorporating the leak
inspection requirements from subpart G into subpart H would
have generated additional confusion in the regulated
community.  The EPA agrees that the leak inspection
requirements which were located in separate sections for each
emission point in subpart G should be condensed into a single
section.  Therefore, in the final rule, the EPA incorporated
all leak inspection provisions for subpart G into- a new
section (§63.148) of subpart G.
     Comment:  One commenter  (A-90-19:  IV-D-73) pointed out
that, although the storage provisions in proposed
§63.120(f)(5) of subpart G indicate that any part of a closed-
vent system subject to subpart H is not subject to subpart G,
this override provision should also be included in
§63.119(e)(1), which specifies the Method 21 monitoring
requirement.
     Response;  As summarized above, the proposed requirements
in §63.120(f) of subpart G for closed-vent systems serving
storage vessels were moved to a new section (§63.148) of
subpart G.  The new provisions in §63.148 have incorporated
the override provision that any part of a closed-vent system
subject to subpart H is not subject to §63.148 of subpart G.
3.5  RECORDKEEPING AND REPORTING
     Comment;  Three commenters (A-90-19:  IV-D-33; IV-D-64;
IV-D-73) contended that the notification requirements for
refilling a vessel after an inspection and for performing a
seal gap measurement were unnecessarily burdensome and should
be omitted from the storage vessel provisions.  One commenter
(A-90-19:   IV-D-32)  made the same comment concerning only the
                              3-45

-------
notification requirement for seal gap measurements.  Two
commenters (A-90-19:  IV-D-33; IV-D-73) reasoned that these
two notification requirements would precipitate_many.  _- .
notifications that would seldom be followed up with a
regulatory agency observer being present.  One commenter
(A-90-19:  IV-D-32) contended that, specifically for seal gap
measurements, it is highly unlikely that a regulatory agency
will have the resources to provide an observer for these
measurements.  One commenter (A-90-19:  IV-D-73)  claimed that
the two notification requirements would merely create an
opportunity for a finding of noncompliance against sources.
     Two commenters (A-90-19:  IV-D-64;  IV-D-73) maintained
that regulatory agencies can request to be notified on a case-
by-case basis.  One commenter (A-90-19:  IV-D-64) suggested
that the HON should allow sources to proceed with the planned
inspections or measurements unless the regulatory agency has
asked to be notified.   The commenter  (A-90-19:  IV-D-64)
remarked that such a provision would protect regulatory
agencies' inspection authority and preserve sources' operating
flexibility.  Two commenters (A-90-19:  IV-D-32; IV-D-33)
suggested that the notification requirements be replaced with
a requirement that the timing and results of the inspections
(A-90-19:  IV-D-33) and measurements  (A-90-19:  IV-D-32;
IV-D-33) be included in the next periodic report.  One
commenter (A-90-19:  IV-D-32) added that facilities could be
required to keep a record on-site of when seal gap
measurements are performed.
     Response;  The EPA has determined that the notification
requirements for an owner or operator to inform the
implementing agency of an upcoming seal gap measurement  (for
EFR vessels) and of vessel refilling when a vessel has been
emptied and degassed (for both IFR vessels and EFR vessels) is
a reasonable requirement that is not unnecessarily burdensome,
as stated by three commenters (A-90-19:  IV-D-33; IV-D-64;
IV-D-73).  These notifications are not required to be
submitted very frequently.  For IFR vessels, which are the
most common type of floating roof vessel used by the SOCMI,
                              3-46

-------
the notification requirement for vessel refilling will be
required once per ten years, or each time the vessel is
emptied and degassed.  For EFR vesselsv the-notification
requirement for vessel refilling has no specified schedule, as
the notification is required each time this type of vessel is
emptied and degassed, according to the schedule established by
the facility operating the vessel.  The EPA anticipates that
EFR vessels will be emptied and degassed no more frequently
than once every ten years.  Also for EFR vessels, the
notification requirement for seal gap measurements will be
required once per year.  The EPA maintains that this
notification requirement is not unnecessarily burdensome, and
that these notifications are necessary for effective
enforcement of the rule.
     The EPA also concluded that these notification
requirements are not likely to result in findings of
noncompliance against sources, as stated by one commenter
(A-90-19:  IV-D-73).  If a source cannot notify the
implementing agency within 30 days due to an unplanned event,
a source is not necessarily in noncompliance.  Both of these
notification provisions specify that if the seal gap
measurement or internal inspection associated with the vessel
refilling were unplanned, then the notification could be made
seven days in advance of the measurement or refilling, rather
than the standard 30 days in advance.
     The EPA has also concluded that, based on discussions
with State agencies, these notifications will result, in many
cases, in observers being sent to facilities to be present
during the measurement or inspection.  The EPA recognizes that
some implementing agencies may choose to send observers to
these measurements and inspections less frequently than other
implementing.agencies; however, the EPA anticipates that the
majority of implementing agencies will use these notifications
for enforcement purposes.
     Comment•  Two commenters (A-90-19:  IV-D-34) (A-90-21:
IV-D-17)  contended that the EPA revise the notification
requirement of a minimum of seven days prior to refilling for
                             3-47

-------
unplanned storage vessel inspections in proposed §63.120(a)(6)
and (b)(10)(iii) to allow for more flexibility.  The
commenters (A-90-19:  IV-D-34) (A-90-21:  IV-D-17) explained
that tanks can be emptied, cleaned, repaired or inspected, and
filled in fewer than seven days,  and for these tanks,  the
seven-day wait would require that the vessel remain out of
service longer than necessary.  One commenter  (A-90-19:
IV-D-34)  claimed that good safety practices require that a
storage vessel be returned to service as quickly as possible
because when a tank is removed from service, special
arrangements must be made for the continued operation of the
production unit and for the distribution of the product.  The
commenter (A-90-19:  IV-D-34) explained that, during these
periods of time, routine operations are modified and mistakes
are more likely to occur.  One commenter (A-90-21:  IV-D-17)
claimed that if a vessel could be inspected and operational in
less than seven days, a source would not want to extend the
down time and costs because of the seven-day wait period, and
would therefore have an incentive to forego the inspection.
     One commenter  (A-90-19:  IV-D-34) suggested replacing the
seven-day notification period prior to refilling with a
notification of the unplanned inspection when the tank goes
out of service, which would include the expected inspection
and refilling dates.  One commenter (A-90-21:  IV-D-17)
suggested adding language that would allow the Administrator
to authorize refilling in less than seven days or allow the
source to forego the delay if the delay would cause
substantial economic loss.  The commenter (A-90-21:  IV-D-17)
offered regulatory language that would specify a notification
requirement for requesting a shorter time for refilling and
explaining why this shorter time was required.
     Response;  The EPA has concluded that the requirement for
a seven-day advance notification of refilling of a storage
vessel for which an inspection was unplanned will not cause
sources to extend the downtime of their storage vessels.
First, the EPA has determined that it will be unusual for
owners or operators to choose to complete the process of
                              3-48

-------
emptying, degassing, cleaning, and inspecting a vessel in less
than seven days.  Second, the storage provisions do not
preclude an implementing agency from negotiating with the
owner or operator to allow a vessel to be refilled earlier
than the seven days, if there is good reason to do so.
Finally, the EPA would like to point out that a seven-day time
period is the minimum time that could be specified in order
for enforcement personnel to arrange to observe the
inspection.
     Comment:  Three commenters (A-90-19:  IV-D-86; IV-D-92;
IV-D-97) contended that the requirement for sources to submit
detailed descriptions of upcoming and past routine maintenance
for storage vessels is unnecessary and burdensome.  The
commenters (A-90-19:  IV-D-86; IV-D-92; IV-D-97) stated that
the proposed storage provisions already include requirements
limiting routine maintenance to a specified number of hours
per year and requiring reporting of the total time that
routine maintenance was actually performed.  Two commenters
(A-90-19:  IV-D-86; IV-D-97) noted that detailed maintenance
plans are not required for other emission points regulated by
the HON.
     Response:  The EPA has included the requirement for
owners or operators of storage vessels to submit detailed
descriptions of past and upcoming routine maintenance for the
storage vessels because of the requirement in the storage
provisions that limits the total number of hours that a
control device may be inoperable due to planned, routine
maintenance.   These detailed descriptions of routine
maintenance are needed for enforcement purposes in order for a
facility inspector to verify what is and what is not routine
maintenance,  in the case that an inspection of a facility
coincides with downtime for a control device.  Furthermore,
this reporting requirement provides the source with some
flexibility to establish a site-specific definition of what
maintenance activities are considered routine maintenance.
Therefore, the EPA has concluded that this reporting
requirement is necessary for enforcement purposes and provides
                             3-49

-------
flexibility to facilities, rather than being unnecessary and
burdensome, as stated by the three conunenters (A-90-19:
IV-D-86; IV-D-92; IV-D-97.)
     Comment;  Two conunenters (A-90-19:  IV-D-73) (A-90-21:
IV-D-17) contended that the 30-day notification requirement in
proposed §63.120(f)(2)(iii)(A) for delay of repair of a
closed-vent system is unnecessary and inconsistent with the
proposed equipment leaks provisions in SS63.171 and 63.182(b)
of the proposed subpart H.  One commenter (A-90-19:  IV-D-73)
recommended that the provision be deleted.  The other
commenter  (A-90-21:   IV-D-17) encouraged the EPA to revise the
notification requirement by allowing the source to include the
necessary  information in the semi-annual report, as required
in the proposed subpart H for delay of repair of various
equipment  leak components.
     Response;  The EPA agrees with the two commenters
(A-90-19:  IV-D-73)  (A-90-21:  IV-D-17) that the 30-day
notification requirement in the proposed §63.120(f)(2)(iii)(A)
for delay  of repair of a closed-vent system is unnecessary and
inconsistent with the proposed subpart H.  The storage vessel
provisions allow a source to delay repair of a closed-vent
system if  the repair would require a process unit shutdown, or
if the emissions of purged material resulting from immediate
repair would be greater than the fugitive emissions likely to
result from delay of repair.  Because the criteria for
eligibility to delay repair are clearly specified in the rule,
the EPA concluded that it is not important that the
implementing agency review a facility's rationale during the
delay of repair.  The EPA has revised these provisions for
closed-vent systems such that, if a source chooses to delay
repair of  a closed-vent system for one of these two reasons,
the source's rationale for this decision may be included in
the periodic report, rather than in a separate notification
requirement.
     Furthermore, the Method 21 inspection provisions  for
closed-vent systems that were included in the storage vessel
provisions have been moved from the storage vessel provisions
                              3-50

-------
of the promulgated rule to the new $63.148 of subpart G.   See
section 3.4.4 of this BID volume for further discussion of
this move.  This new §63.148 requires a source to  include  the
specified information concerning the repair extension in the
same periodic reports required by the equipment  leaks
provisions in subpart H.
     Coiiflflefl't;  One commenter (A-90-19:  IV-D-33) requested
that the EPA clarify the wording in proposed $63.120(a)(4),
(b)(7)(ii), and  (b)(8) of subpart G which specifies how
sources may request up to two extensions of up to
30 additional days each for repairing floating roof failures
discovered during inspections or seal gap measurements.  The
commenter (A-90-19:  IV-D-33) pointed out that the provisions
do not specify an approval deadline by the Agency, what
criteria the Agency will use to review the request, whether or
not the Agency will provide a written response to the request,
or whether the source can proceed under the requested
extension without an approval notification.  The commenter
(A-90-19:  IV-D-33) stated that without these details
clarifying the responsibilities of both the regulated facility
and the regulatory authority, the source is uncertain as to
the amount of time available to complete repairs.  The
commenter (A-90-19:  IV-D-33) recommended adding language  to
S63.122(h)(3) to specify that the source will automatically
receive the requested extension unless the Administrator
denies the extension within 15 days of receiving the request.
     Response;  In considering the comment, the  EPA determined
that the requirements specified in proposed SS63.120(a)(4),
(b)(7)(ii), and  (b)(8) of the proposed storage vessel
provisions, requiring sources to request up to two 30-day
extensions for repair of floating roofs, create  an excessive
burden for the implementing agency.  The EPA does not expect
that implementing agencies will have the resources to review
and approve these requests before the requested  time has
elapsed,  i.e., 30 days.  The EPA concluded that  sources that
use up to two 30-day extensions will be required to report the
decision to use each 30-day extension in the next periodic
                             3-51

-------
report.  This report shall include the same information that
was required, in the proposed storage provisions, to be
included in the request for the extensions  (1) a description
of the failure; (2) documentation that alternate storage
capacity is unavailable; and (3) a schedule of actions that
the source will take in order to repair the control equipment
or empty the storage vessel as soon as possible within the
30-day extension period.
3.6  WORDING OF THE PROVISIONS
     Comment:  Two commenters (A-90-19:  IV-D-64) (A-90-21:
IV-D-17) stated that the wording of S63.120(f)(1), describing
the proposed frequency of inspecting closed-vent systems for
leaks, is ambiguous.  The commenters (A-90-19:  IV-D-64)
(A-90-21:  IV-D-17) explained that the wording could be
misinterpreted to mean that inspections must be done during
all filling operations or just during initial filling, in
addition to once per year when the storage vessel is not being
filled.  One commenter (A-90-19:  IV-D-64) suggested excluding
the wording "during filling of the vessel."  The other
commenter (A-90-21:  IV-D-17) suggested revising the language
of §63.120(f)(1) to read: "Inspections of the closed-vent
system shall be done at least once per year while the vessel
is being filled.1*
     Response:  The EPA would like to clarify that the phrase
"during filling" was not intended to mean that inspection was
required each time a vessel was filled.  Rather, it was
intended to require inspection during worst case conditions.
While a tank is being filled, the flow rate and pressure of
the emission stream are at their highest.  Thus, a failure in
the closed-vent system will be more easily detected.  However,
because the proposed language was confusing to some
commenters, the EPA concluded that the wording of proposed
S63.120(f)(1) should be modified to more clearly reflect the
EPA's intended meaning.  The requirement is in $63.120(d)(6)
of the final rule, and the revised language is:  "The  initial
and annual inspections...shall be done during  filling  of the
storage vessel."
                             3-52

-------
               One commenter  (A-90-19:  IV-D-64) suggested
that the language in 'proposed $63. 122 (d) which describes the
reporting requirements for inspections of IFR vessels is
unclear.  The commenter  (A-90-19:  IV-D-64) explained that the
proposed language could be interpreted as requiring the
reporting of inspection results for all storage vessels
inspected, including storage vessels with no detected
failures, in the event that a single storage vessel failed its
inspection.  The commenter (A-90-19:  IV-D-64) suggested
alternative wording.
     Response:  The EPA reviewed the language in proposed
S63.122(d) and concluded that the wording in the paragraph
does not imply that all inspected storage vessels at a source,
regardless of defects, would require reporting.  The EPA
determined that the language in §63. 122 (d) clearly refers to a
single inspection rather than to a group of inspections
(i.e., the whole event of inspecting all storage vessels with
the given control equipment).  In proposed S63.122(d), the
regulation refers to "each inspection conducted in accordance
with S63.120(a),H and proposed $63. 120 (a) refers to only a
single storage vessel and never to a group of storage vessels.
Therefore, the wording of proposed $63. 122 (d) will remain the
same in the final rule.
               One commenter (A-90-19:  IV-D-34) stated that
the recordkeeping requirement in proposed $63. 123 (a) that
records be maintained of each Group l and Group 2 storage
vessel's dimensions and of an analysis showing its capacity,
as long as the vessel is in service, needs to be reworded to
clarify that these records must be maintained only if the
vessel is in organic HAP service.  The commenter  (A-90-19:
IV-D-34) explained that the service of a storage vessel may
change several times over its lifetime, depending on the needs
of the facility.
     Response:  The EPA agrees that the language in $63. 123 (a)
should be clarified as suggested by the commenter (A-90-19:
IV-D-34) .  The EPA changed the wording to read: "This record
shall be kept as long as the storage vessel retains Group 1 or
                             3-53

-------
Group 2 status and is in operation."  A storage vessel that
retains Group 1 or Group 2 status is a storage vessel that is
covered by the HON and, thus, is in organic HAP service.  If a
storage vessel discontinues organic HAP service, it no longer
retains Group 1 or Group 2 status.
     Comment;  One conunenter (A-90-19:  IV-D-87) reported that
proposed §63.120(b)(3) of subpart G has a typographical error
in that 212 square inches should be replaced with 21.2 square
inches.
     Response;  It is not clear to the EPA why the conunenter
is suggesting that there is a typographical error in
§63.120(b)(3).  Because the commenter is suggesting that the
accumulated area of gaps per meter of vessel diameter for the
primary seal be changed to 21.2, which is the maximum allowed
area for the secondary seal in §63.120(b)(4), the EPA has
interpreted that the commenter assumed the requirements for
primary and secondary seals to be the same.  The EPA would
also like to point out that the commenter's suggested change
should have been presented as 21.2 square centimeters rather
than square inches.  The EPA wishes to clarify for the
commenter that the seal gap requirements are different for
primary and secondary seals.  For the primary seal, the
storage provisions specify an allowance of up to 212 square
centimeters of accumulated area of gaps per meter of vessel
diameter.  For the secondary seal, the storage provisions
specify a more stringent allowance of up to 21.2 square
centimeters of accumulated area of gaps per meter of vessel
diameter.  The requirement for a secondary seal is more
stringent than that for a primary seal because a secondary
seal is exposed directly to the atmosphere.  A gap in the
secondary seal has a greater potential to cause organic HAP
emissions than does the primary seal because the primary seal
in a two-seal system is protected from wind effects by the
presence of the secondary seal.  These requirements are the
same requirements specified in 40 CFR part 60 subpart Kb, the
NSPS for Volatile Organic Liquid storage Vessels, and in
                              3-54

-------
40 CFR part 61 subpart Y, the NESHAP for Benzene Emissions
from Benzene Storage Vessels.
3.7  MISCELLANEOUS
     comment:  One commenter (A-90-19:  IV-D-92) requested
clarification of the meaning of the term "impurity", as used
in the storage vessel provisions.
     Response:  The EPA agrees with the commenter  (A-90-19:
IV-D-92) and has revised the definition of the term
"impurity".  The revised definition in §63.101 of  subpart F
reads as follows:  "a substance that is produced
coincidentally with the primary product, or is present in a
raw material.  An impurity does not serve a useful purpose in
the production or use of the primary product and is not
isolated."
     Comment;  one commenter (A-90-19:  IV-D-92) requested
further clarification of the definition of the term "product",
as used in the storage vessel provisions.  The commenter
(A-90-19:  IV-D-92) explained that it is difficult to identify
the "product" for mixed streams, such as gasoline.
     Response;  The EPA would like to clarify for the
commenter (A-90-19:  IV-D-92) that the term "product", as used
in the storage provisions of the rule, refers to the material
stored in the storage vessel.  For mixed streams,  such as
gasoline, identifying the "product" is the same as identifying
the mixed streams included in the gasoline.
     Comment?  One commenter (A-90-19:  IV-D-92) stated that
it is important that the HON be consistent with other
regulations and asked if the HON is consistent with
40 CFR 112, The Oil Pollution Act.
     Response:  Although the HON and the Oil Pollution
Prevention regulation (40 CFR 112) have different objectives
and are unlikely to regulate the same facilities or
pollutants, their provisions are not inconsistent.  The Oil
Pollution Prevention regulation is written to prevent oil
discharges into or on navigable waters from facilities
engaging in all aspects of acquiring and selling oil and oil
products, including storage.  The Oil Pollution Prevention
                             3-55

-------
regulation (1) requires that a spill prevention control and
counter-measure plan be written and implemented; and
(2) provides applicable guidelines for prevention of leaks,
such as adequate design, secondary containment, liquid level
sensors, and periodic inspections and tests.  The HON
requirements are designed to limit emissions of HAP vapor from
storage vessels, rather than to prevent leaks and spills;
however, none of the HON provisions preclude spill plans,
secondary containment, inspections, etc., required or
suggested in 40 CFR 112.  Therefore, a facility could comply
with both regulations.
                              3-56

-------
                    4.0  TRANSFER OPERATIONS

4.1  APPLICABILITY AND GROUP I/GROUP 2 DETERMINATION
     Comment;  Two commenters  (A-90-19:  IV-D-32; IV-D-112)
concurred that the EPA had appropriately defined RCT for
transfer operations.  Two commenters (A-90-19:  IV-D-58;
IV-D-62) supported the EPA's decision to propose the floor
level of control as MACT for transfer operations.  The
commenters  (A-90-19:  IV-D-58; IV-D-62) stated that the
increased cost of controlling all racks is not justified given
the very small emissions reductions.  One commenter (A-90-19:
IV-D-92) stated that the HON transfer operations provisions
appear to be consistent with other NESHAP's.
     Response:  The EPA appreciates the commenters support.
     Comment:  Several commenters (A-90-19:  IV-D-32;
IV-D-112) supported vapor balancing being included in the HON
as an option for emissions averaging.  The commenters
(A-90-19:  IV-D-32; IV-D-112) also supported the exclusion of
racks using vapor balancing from the transfer operation
provisions.
     Response:  The EPA appreciates the commenters support.
     Comment:  One commenter (A-90-19:   IV-D-85)  maintained
that in order to escape the conclusion that the floor levels
require control, the EPA divided the transfer racks into two
groups, those with average vapor pressures above or below
1.5 psia.
     Response;  It is not practical or cost effective to
control all emission points at a facility.  Emissions from
some points are very small.  The cost to control these
emissions is relatively large,  and the quantity of emissions
reduced does not warrant the cost of controlling them.   This
                              4-1

-------
is a fact recognized by State environmental agencies and
evident in applicability criteria found in existing
regulations.  In the case of transfer operations, the* current
regulations generally define applicability based art .vapor
pressure and throughput.  These two factors have the greatest
effect on the magnitude of emissions from transfer operations.
Host State regulations require control of transfer operations
only at liquid vapor pressures above 1.5 psia.  Therefore, for
the HON analysis, it was a natural division of data; racks
transferring chemicals with a vapor pressure less than
1.5 psia have low emissions and are not controlled; racks
transferring chemicals with a vapor pressure greater than
1.5 psia have larger emissions and, therefore, require
control.
     Comment:  One commenter (A-90-19:  IV-D-85) asserted that
the EPA's determination of the floor for new sources is
inconsistent with the data presented in a memorandum in the
docket.  The commenter  (A-90-19:  IV-D-85) contended that the
memorandum shows that racks with vapor pressures less than
1.5 psia are sometimes controlled and the best-controlled rack
from a volume standpoint is controlled with a throughput of
160,000 gallons/yr.
     Response:  The commenter is correct in noting that
attachment 2 of the docket memorandum "Analysis of Hazardous
Organic NESHAP (HON) Database to Determine the Floors,"
(Docket item A-90-19:  II-B-277) indicates that two racks
identified during floor determination with average rack
weighted vapor pressure less than 1.5 psia are controlled.
However, the information for these two racks is not an
accurate representation of what actually occurs at SOCMI
facilities, but is an artifact of the data assumptions used in
the analysis.  These racks were assigned control because there
are some chemicals loaded at the racks that have a vapor
pressure greater than 1.5 psia, and, therefore, are required
by State regulators to be controlled.  In reality, transfers
of the chemicals with a vapor pressure below 1.5 psia would
                              4-2

-------
not be controlled but transfer of the chemicals with a vapor
pressure above 1.5 psia would be.
     The commenter is also correct in noting that the lowest
throughput for controlled racks with vapor pressures greater
than 1.5 psia is 160,000 gallons/year instead of
170,000 gallons/year.  The EPA decided to make the
applicability criteria for new and existing sources the same
for simplicity.  The actual difference between the two levels
(10,000 gallons/year) is about the same as the capacity of one
tank truck.  Also, there will be very few facilities with a
rack at a new source with a throughput greater than
160,000 gallons/year and less than 170,000 gallons/year.  The
EPA expects that this change would have only a minimal effect
on emissions.  Therefore, it was not incorporated into the HON
transfer provisions.
     Comment:  One commenter (A-90-19:  IV-D-32) contended
that the EPA has not demonstrated that existing control
devices on Group 1 transfer racks can actually achieve
98-percent control efficiency of HAP's.  The commenter
(A-90-19:  IV-D-32) recommended that the EPA either verify
that 98-percent control of HAP's is achievable for existing
control devices or evaluate the retrofit costs and the
incremental benefit to the environment for facilities that
must remove existing control devices achieving 95 percent and
replace them with slightly more efficient control devices.
     Response;  The 98-percent control is based on studies
used to determine VOC control levels for past NSPS and has not
been proven by testing for each individual HAP.  These two
issues do not weaken the EPA's decision for 98-percent control
of HAP's for the following reasons:  (1)  nearly all HAP's are
also VOC's; and (2) HON compliance is not based upon control
of each individual HAP.   Compliance with the HON may be based
upon measurements of either total organic HAP or TOC.
Clearly, a control device might have a higher level of control
for one particular HAP than for another,  but compliance is
based on the overall reduction of total organic HAP or TOC
from an emission point.
                              4-3

-------
     The 98-percent level of control was chosen because it has
been shown to be uniformly achievable by well-designed and
operated combustion devices.  Test data to demonstrate
efficiency in a thermal, incinerator are not available for each
individual HAP.  However, the efficiency conclusions for a
thermal incinerator (98-percent reduction or an outlet
concentration of 20 ppmv) were based on test data using the
most difficult VOC compounds to combust, which included
several organic HAP's.  Therefore, it was concluded that the
98-percent reduction can be achieved for total organic HAP for
all well-designed and operated systems  (A-79-32, II-B-31).
The EPA recognizes that thermal incineration may achieve
greater than 98-percent reduction, but test data show that
levels greater than 98 percent may not be uniformly achievable
under all operating conditions.
     The commenter is correct that some existing control
devices may not be achieving 98-percent control of HAP's and
may have to be retrofitted or replaced in order to meet the
requirements of the HON transfer operations provisions.  The
EPA considered costs to these facilities by applying a flare
or incinerator to the outlet stream of any existing control
device achieving less than 98-percent HAP reduction.
Therefore, the estimation of incremental benefits does address
the additional costs to these facilities.
     Comment:  One commenter (A-90-19:  IV-D-41)- contended
that the EPA did not identify the control level achieved by
the top 12% of transfer racks.   The commenter (A-90-19:
IV-D-41) also stated that the EPA provided no analysis of how
many facilities have vapor balancing and vapor collection,
although these were identified as superior control techniques
since transfer racks using this technique are exempt from the
HON.  The commenter (A-90-19:  IV-D-41) questioned why vapor
balancing and vapor collection were not identified as the
floor or MACT.
     One commenter (A-90-19:  IV-D-85) noted that the Texas
new source review program requires control of transfer racks.
Hence, the commenter (A-90-19:   IV-D-85) asserted that the
                              4-4

-------
floor level of control should take into account those
facilities for which a BACT or LAER determination has been
made prior to June 31, 1991.
     Response:  The top 12% of the.-transfer Bracks achieve 98%
reduction [see memorandum titled "Analysis of Hazardous
Organic NESHAP (HON) Database to Determine the Floors," Docket
item A-90-19:  IV-B-277].  Using vapor balancing with vapor
collection on a transfer rack exempts the facility from the
HON transfer provisions because the EPA's technical analysis
showed that under typical conditions, vapor balancing reduces
emissions by 98 percent or better, (see memorandum titled:
"Efficiency of a Vapor Balancing System," Docket item A-90-21:
II-B-28).  However, data were not available to identify which
SOCMI facilities used vapor balance to control emissions.
     Due to the strict schedule for the HON to be proposed and
promulgated, time was not available to survey SOCMI facilities
or BACT and LAER information.  In addition, the EPA is not
required by the Act to survey all SOCMI facilities.  The Act
requires the EPA to set the MACT floor for the best-performing
12 percent of existing sources for which the EPA has emissions
information.  The EPA considers a review of State and Federal
regulations to adequately characterize the controls achieved
for the SOCMI nationwide.  Information supplied by commenters
can also be included in the MACT floor analysis.  However, the
commenter did not provide any information on how widely vapor
balancing is used.
4.1.1  Applicability
     Comment;  Three commenters (A-90-19:  IV-D-58; IV-D-63;
IV-D-73) requested that the EPA clarify that the HON is
applicable only to transfers of materials from SOCMI
processes, and not all materials transferred at any given rack
or arm.   Two commenters (A-90-19:   IV-D-58; IV-D-63)  suggested
wording for changing the "transfer operation" definition to be
applicable only to a major source facilities associated with
the transfer of product for one or more chemical manufacturing
processes specified in §63.100.   Another commenter (A-90-19:
IV-D-64) suggested that the definitions of "Group 1 transfer
                              4-5

-------
racks" and "rack-weighted average vapor pressure1? specify that
they do not apply to non^SOCMI -product transfers.
     Response;  The transfer provisions apply to transfer
racks as defined "in "subpart'-F,-when used for-transfer-of
HAP's.  It is possible that HAP's associated with chemical
manufacturing process units not subject to the HON will be
subject to control under the HON if they are transferred at a
transfer rack.  The definition for "transfer rack" has been
revised to clarify this.  See section 4.1.2 for further
explanation.
     The EPA has replaced the "rack-weighted average vapor
pressure" definition with a definition for "rack-weighted
average partial pressure" as discussed in the response to the
next comment.
     Comment;  One commenter (A-90-19:  IV-G-4) contended that
the definition of "impurity" in §63.101 of the proposed HON
combined with the definition for "Group 1 transfer rack" and
"rack-weighted average vapor pressure" in §63.111 of the
proposed HON, and the calculation techniques in §63.130(g)(3)
of the proposed HON will require control of low HAP content
streams.  The commenter  (A-90-19:  IV-G-4) asserted that
facilities will be forced to place controls on transfer
operations that load materials containing small amounts of
HAP's.
     The commenter (A-90-19:  IV-G-4) suggested that the
definition of impurity specify that HAP's considered as
impurities not be intended to be part of the product being
processed, stored, or transferred.  The commenter (A-90-19:
IV-G-4) also suggested that the rack-weighted average vapor
pressure be defined as the organic HAP's partial pressure
considering all materials loaded at the particular loading arm
or station.  The commenter  (A-90-19:  IV-G-4) requested that
the definition of rack-weighted average vapor pressure or
§63.130(g)(3) in the proposed HON include an equation for the
calculation of rack-weighted average vapor pressure.
     Response;  The EPA agrees with the commenter's suggestion
for defining an organic HAP rack-weighted average partial
                              4-6

-------
pressure instead of a HAP rack-weighted average vapor
pressure.  The final transfer operations provisions have been
revised to replace the definition of rack-weighted average
vapor pressure with a definition on rack-weighted partial
pressure.  The definition is as follows:
     Rack-weighted average partial pressure means the
     throughput weighted average of the average maximum
     true vapor pressure of liquids containing organic
     HAP's transferred at a transfer rack.  The rack-
     weighted average partial pressure shall be
     calculated using the equation below:

                              2 Pi Gi
                          P = - i — i
where :
     P    =    Rack-weighted average partial pressure,
               kilopascals
     Pi   =    Individual HAP maximum true vapor pressure,
               kilopascals
     Gi   =    Yearly volume of individual organic HAP
               transferred at the rack, liters
Also, all requirements referring to rack-weighted vapor
pressure have been changed to rack-weighted partial pressure.
Because of this change, facilities will only be required to
control racks if there is sufficient HAP to make the rack a
Group 1 transfer rack (i.e., loads greater than 0.65 million
liters per year of liquid products containing organic HAP's,
and with a rack-weighted HAP partial pressure of 10.3 kPa or
greater) .
     The definition of impurity already implies that it is not
"intended" to be part of the product by stating an impurity is
"produced coincidentally. "   The EPA considers this definition
to be clear.
     Comment :   Two commenters (A-90-19:  IV-D-64 ;  IV-D-73)
requested that the EPA clarify that the provisions in the HON
refer to the reduction of the total HAP content and not to
each individual HAP.   One commenter (A-90-19:  IV-D-64)
specifically suggested that §§63. 126 (b) (1) ,  63. 128 (a) (7) ,  and
                              4-7

-------
63.129(a)(4) in the proposed HON be edited by the addition of
the word "total" so that they refer to total organic compound
concentration and total HAP concentration.
     Response; . The EPA agrees.,with the commenters that total
HAP concentration should be referred to in the transfer
provisions, and these changes have been made in the final
transfer operations provisions.  The term "total organic
compound concentration" is used in the process vents
provisions to refer to compounds measured according to the
procedures of Method 18 of 40 CFR part 60, appendix A.
Therefore,  the term "total organic compound concentration"
cannot be used in the transfer operations provisions where an
organic compound concentration could be measured using
Method 18 or Method 25A.  The EPA considers the term "organic
compound concentration" adequate to imply a total organic
compound concentration.
4.1.2  Group l/Group 2 Determination
     Comment:  One commenter (A-90-19:  IV-D-32) supported the
selection of throughput and vapor pressure as reasonable
criteria for defining Group 1 transfer racks.  However, three
commenters  (A-90-19:  IV-D-32; IV-D-73; IV-G-4) contended that
the EPA should clarify whether the provisions refer to
transfer racks or arms.  One commenter (A-90-19:  IV-D-73)
stated that the definitions for "Group 1" and "rack-weighted
vapor pressure transfer rack" do not address a loading arm
criterion as in §63.100(b)(5) in the proposed HON.  The
commenter  (A-90-19:  IV-D-73) recommended adding definitions
for "Group 1 transfer arm" and "arm-weighted vapor pressure."
Another commenter  (A-90-19:  IV-G-4) recommended that the
phrase "loading arms and associated equipment dedicated to
specific filling operations" in §63.110(d)(1)(i) and
(d)(2)(ii)  in the proposed HON be substituted for racks; and
in §§63.110(d)(2)(i  and  (d)(2)(ii) in the proposed HON the
phrase "filling operations in non-dedicated loading arms and
associated equipment" be substituted for "operations."  One
commenter  (A-90-19:  IV-D-64) asserted that the references to
                              4-8

-------
loading racks should be changed to refer to loading arms
because a loading rack may transfer non-SOCMI chemicals.
     Response;  The EPA has clarified many of the
applicability and group determination provisions regarding
transfer operations in the proposed HON.  The intent of these
provisions has not changed.
     The intent of the language in proposed §63.100(b)(5) of
subpart F and §63.100(h) of subpart F in the final rule is to
assign loading racks, loading arms, or loading hoses to a
chemical manufacturing process unit.  If the chemical
manufacturing process unit is~subject to the HON, then the
loading rack, arm, or hose is also subject to the HON.
     Once it is determined that a rack, arm, or hose is
subject to the HON, group status must be determined.  Group
status is determined for a transfer rack.  A transfer rack is
defined as the collection of all arms or hoses that are
assigned to a chemical manufacturing process unit that is
subject to the HON.  For example, if a facility has a rack
that consists of eight arms and six of these arms are assigned
to chemical manufacturing process units subject to the HON and
the other two load petroleum refinery products not subject to
the HON, then the "transfer rack", as defined for the HON, is
made up of the six arms that are assigned to the chemical
manufacturing process unit subject to the HON.  Group status
must be determined based on the vapor pressures and
throughputs of the HAP's loaded at the arms at a rack that are
subject to the HON.  In cases where a rack or arm(s) has been
assigned to a chemical manufacturing process unit subject to
the HON, the rack or collection of arms must be controlled
during transfers of all HAP's regardless of whether those
HAP's were associated with SOCMI chemical manufacturing
process units.
     The EPA determined that assigning equipment to be subject
to the HON would be easier from an enforcement and control
perspective, as opposed to assigning transfer operations.
Also, by assigning equipment, applicability will be more
                              4-9

-------
obvious when other rules are promulgated affecting transfer
operations under other source categories.
     Comment:   Three commenters (A-90-19:  IV-D-32; IV-D-112,
IV-G-4) requested that the EPA clarify -the calculation of an
annual rack-weighted average HAP vapor pressure.  One
commenter (A-90-19:  IV-D-34) noted that the definition of
rack-weighted vapor pressure does not define the temperature
to be used in the calculation or specify a method to be used
if materials are loaded at different temperatures.  Two
commenters (A-90-19:  IV-D-34; IV-D-112) suggested that the
EPA modify the definition of rack-weighted vapor pressure to
specify that the vapor pressure should be calculated for each
material at the maximum average monthly temperature of the
materials loaded.  One commenter (A-90-19:  IV-D-32) added
that loading vapor pressure should be established at the
annual average temperature of the material loaded.
     Response:  The commenters are correct that the proposed
transfer operations provisions did not specify the temperature
to be used in the calculation of the annual rack-weighted
average HAP vapor pressure.  The provisions have been changed
to specify that the maximum true vapor pressure be used in the
calculation for the vapor pressure.  The definition of maximum
true vapor pressure has been edited to make it appropriate for
the transfer calculation.
     Also, it has been specified in the emissions averaging
provisions §63.150(g)(4) that the temperature to be used is
the annual average loading temperature for a given chemical.
Specifications have also been added to the emissions averaging
transfer calculation provisions in order to handle the
calculation of credits/debits for racks that transfer liquids
at different temperatures.
     Comment:   One commenter (A-90-19:  IV-D-33) recommended
that Group 2 transfer racks that infrequently load materials
with vapor pressures greater than 10.3 kPa should not be
required to calculate rack-weighted vapor pressure.  The
commenter (A-90-19:  IV-D-33) stated that if a plant routinely
uses a Group 2 transfer rack as a dedicated rack to an organic
                             4-10

-------
product with HAP vapor pressures less than 10.3 kPa, the plant
would probably not monitor a precise vapor pressure record as
long as it could document that it has less than 10.3 kPa.
     Response:  The requirement for calculating the average
rack-weighted partial pressure (the vapor pressure was changed
to partial pressure in the final rule) is not overly
burdensome, and it would not be additionally burdensome in the
situation described by the commenter.  It is not necessary to
"monitor a precise vapor pressure."  The chemical transferred
can simply be recorded and then the maximum true vapor
pressure can be calculated.  The maximum true vapor pressure
is also needed under the storage vessel provisions so that it
is readily available.
4.2  COMPLIANCE
     Comment;  Two commenters (A-90-19:  IV-D-90; IV-D-100)
asserted that the compliance options for transfer operations
using vapor balancing are crude and difficult to achieve.
     Response:  No details were provided by the commenter as
to why the vapor balancing compliance options are crude and
difficult to achieve.  Vapor balancing is frequently used in
industry as an emission control and as a product recovery
technique.  A technical analysis showed that under typical
conditions, vapor balancing reduces emissions by 98 percent or
better (see memorandum titled:  "Efficiency of a Vapor
Balancing System," Docket item A-90-21:  II-B-28).  Therefore,
if facilities meet the definition of vapor balancing, they are
considered to be in compliance with the provisions.
4.2.1  Performance Testing
     Comment:  One commenter (A-90-19:  IV-D-73) supported the
exclusions in §63.128(c)  from performance test requirements,
particularly the exclusion for combustion with primary fuel in
a boiler or heater.
     Response;   The EPA appreciates the commenter's support.
     Comment:  One commenter (A-90-19:  IV-D-64) stated that
the span value for Method 25A in the HON transfer provisions
should be allowed to be between 1.5 and 2.5 times the
concentration measured, as it states in the method, rather
                             4-11

-------
than exactly twice the concentration measured, as specified in
the HON.
     Response;  The EPA agrees with the commenter, and has
changed the regulation.  The span value is a derived number
based on an assumption of the concentration anticipated to be
measured.  The calibration gas concentrations needed are then
determined based on percentage ranges of this span value.
Allowing a range for the span value (and hence calibration
gases) facilitates meeting these criteria by potentially
allowing the same calibration gases to be used for different
tests while not adversely affecting the objective of having
the measured concentration bracketed by two of the calibration
gases.
     Comment:  One commenter (A-90-19:  IV-D-35)  guestioned
which HAP should be used as the density term in the equation
used to calculate the mass of organic compounds emitted during
a testing interval.  The commenter (A-90-19:  IV-D-35) also
questioned why the organic compound concentration corrected to
3 percent oxygen was not used in the calculation instead of
the total concentration of organic compounds.
     The commenter (A-90-19:  IV-D-35) recommended that the
equation for determining the mass of organic compounds emitted
during each testing interval use the concentration of each
organic compound emitted during the testing interval  (Cjn);
and be represented by the following equation:

                       Mj = FVs
                                n=l

     Response:  Method 25A, for measurement of Cjn, does not
speciate the HAP content of the stream.  The intent of
measuring the mass of organic compounds is for calculation the
percent reduction across a control device.  In the step where
the percent reduction across the control device is calculated,
the density cancels out and is not needed.  The density term
is only provided in the equation so that, in the calculation
of the mass of organic compounds, units cancel out.  The
                             4-12

-------
 transfer operations provisions have been revised to include
 the density of hexane in the calculation of percent reduction
 in order to maintain consistent units; however, the value of
 the density used has no bearing on the percent reduction
 calculated.
      As in the case of density, a correction of the organic
 concentration to 3 percent oxygen does not affect the percent
 reduction calculated.  The oxygen correction is only necessary
 for determining compliance with the 20 ppmv exit concentration
 requirement.
      Comment;  One commenter (A-90-19:  IV-D-69) claimed that
 performance tests over one loading cycle are sufficient
 because filling times for large vessels may take several
 hours, and three multiple loadings may not typically occur at
 one time.  Another commenter (A-90-22:  IV-D-13) contended
 that it is unduly burdensome to require a performance test
 duration of three loading cycles for infrequent loading
 operations.
      One commenter (A-90-22:  IV-D-13) stated that, for small-
 volume loading operation owners and operators who have to hire
 contractor help, a significant portion of the costs of
 conducting a three-loading-cycle performance test would be for
 setting up and breaking down equipment and additional
• transportation and living expenses (especially if they are
 located in a distant city, or out of state), in addition to
 the monitoring and analytical work.  The commenter (A-90-22:
 IV-D-13)  suggested that the EPA provide an option that
 infrequent loading operations (once a week or less) may be
 performance tested for only one loading cycle with duplicate
 samples.
      Response:  Typically, performance tests are the average
 of three runs and,  because of the variability in the
 concentration of HAP's over a loading cycle, a run was defined
 as a loading cycle for transfer operations.  The original
 intent of testing three different loading cycles was to
 acquire sufficient data to ensure that a control device is
 operating properly.   However, the EPA does consider this
                              4-13

-------
requirement overly burdensome for sources that transfer small
amounts of organic HAP's.  It is possible for a transfer rack
to be considered a Group 1 rack and only transfer into 17 tank
trucks or 9 rail cars per year.  Requiring testing during
three loadings could conceivably be spread out over months.
In order to reduce the burden to these sources, the final
transfer provisions include an exception for racks that
transfer less than 11.8 million liters per year.  For these
racks, an initial performance test is not required if a design
evaluation is provided.  The permit authority can require a
performance test any time there is a concern that the control
device is not working properly.
     The suggestion by the commenters to require that only one
loading cycle be tested for the performance test was
considered; however, the EPA maintains that a test of one
loading cycle could not provide sufficient data on the
performance of the control device, and therefore would not be
meaningful.  The EPA opted to require a design evaluation in
these situations where sources load less than 11.8 million
liters per year through a transfer rack.
     A source is required to document that the control device
used achieves the required control efficiency during
reasonably expected maximum loading conditions.  Monitoring
parameters can be determined using engineering judgement, or
alternatively, a performance test can be conducted to
determine monitoring parameters.  For carbon adsorbers and
condensers, the parameters are provided in §63.128(h) of the
final rule.
     Comment:  One commenter (A-90-22:  IV-D-13) contended
that the EPA should revise §63.128(a)(8)(ii)(A) regarding the
emission testing interval requirements to ensure consistency
with §§63.129(a)(4)(iii) and 63.129(a)(6)(iii).  The commenter
(A-90-22:  IV-D-13) contended that in various sections of the
proposed transfer provisions 5-minute and 15-minute testing
intervals are required but in §63.128(a)(8)(ii)(A) only
5-minute testing intervals are allowed.
                             4-14

-------
     Another commenter  (A-90-19:  IV-D-64) asserted that
15-minute recording intervals are adequate for performance
tests.  The commenter (A-90-19:  IV-D-64) asserted that
requiring 5 minute recording intervals for performance tests
shorter than 3 hours was unnecessary.  The commenter  (A-90-19:
IV-D-64) suggested that the recording frequency be "no less
frequently than every 15 minutes" in §63 .129 (a) (4) (iii) ,
§63.129(a)(6)(iii), §63.130(a)(1)(i) and §63.130(a)(1)(ii) to
allow flexibility to monitor more frequently when it  is in the
source's interest to do so.
     Response:  The EPA agrees with the commenters that
5-minute testing intervals are not necessary.  The EPA has
determined that 15-minute testing intervals should adequately
identify performance that is outside of the approved  operating
parameter ranges.  The EPA has also added language to the
transfer provisions for recording data "at least every
15 minutes".
     Comment:  One commenter (A-90-19:  IV-D-77)  supported the
use of either Method 18 or Method 25A for measuring vent
stream HAP content from transfer rack applications.   However,
the commenter (A-90-19:   IV-D-77) contended that obtaining
certified calibration standards for all the HAP's may not be
possible.
     Response;  The EPA agrees that gaseous standards are not
commercially available for all the HAP's.  However,
section 6.2 of Method 18 allows the option of preparing
gaseous standards either from a higher concentration gas
cylinder or through liquid or gas injection and provides a
procedure for preparing the standards.
     Comment;  One commenter (A-90-19:  IV-D-73)  related that
many of their plants have a large number of simultaneous
fillings of SOCMI and non-SOCMI materials venting to common
control devices and loading would have to be suspended for
days or weeks to individually evaluate each covered transfer
operation.   The commenter (A-90-19:  IV-D-73) requested that
the EPA revise the test methods section for transfer
operations to reflect the use of shared control devices for
                             4-15

-------
multiple simultaneous loading operations and for loading,
processes, and other operations.  The commenter (A-90-19:
IV-D-73) recommended that an approach similar to §63.116(c) be
used which calls for testing at the inlet and outlet of the
control device, where performance testing is required.  The
commenter (A-90-19:  IV-D-73) stated that the test
requirements must be uncoupled from the loading cycles to make
this feasible for large loading operations with common control
devices.
     The commenter (A-90-19:  IV-D-73) also stated that the
definition of loading cycle should be revised to refer to the
time at which flow to a transfer operation control device
begins to the time it ends, regardless of how many
simultaneous or overlapping loadings are occurring.
     Response:  Although the proposed rule did not preclude
shared control devices, it was not clear how an owner or
operator would demonstrate compliance.  Provisions have been
added to the rule to clarify that in case of a control device
being shared with process vents, the performance test required
under the process vent provisions is adequate for the transfer
provisions.  For control devices shared among transfer racks
or arms where materials are simultaneously loaded, the
performance test requirements have been revised to be similar
to the process vent provisions:  three one-hour performance
test runs.  Simultaneous loading occurs when the beginning and
ending times of loading cycles coincide and overlap such that
there is no interruption in vapor flow to the control device;
as one loading cycle is completed, another one begins or has
already begun.
     Comment;  Three commenters (A-90-19:  IV-D-13; IV-D-64;
IV-D-73) discussed issues regarding the compliance
determination for  flares.  Two commenters (A-90-19:  IV-D-64;
IV-D-73) contended that the integrated sampling to measure
flow rate, required in §63.128(b)(1)(ii) was unnecessary.  One
commenter (A-90-19:  IV-D-64) contended that times when a
compliance determination must be made are not clear.  The
commenter (A-90-19:  IV-D-64) asserted that the intent of the
                             4-16

-------
provisions was for an initial performance test to determine
compliance with the flare provisions in §63.11(b) of the
General Provisions and not to require a compliance
determination each time HAP's are loaded and controlled by a
flare.
     One commenter (A-90-22:  IV-D-13) suggested that the EPA
modify the 2-hour observation period requirements when a
loading period is less than 2 hours for flare performance
tests.  The commenter (A-90-22:  IV-D-13) recommended
modifying the provisions to allow either 2 hours or until the
loading cycle is completed, whichever is less.
     Response:  The intent of the provisions, as the commenter
stated, are to require a performance test for all control
devices other than flares and a compliance determination for
flares.  The requirement that performance tests are initial is
stated in §63.103(b)  of subpart F of the proposed rule.
Section 63.103(b)(l)  of subpart F of the proposed rule refers
to §63.7(a) of the General Provisions for the schedule and
procedures for performance tests.  Section 63.103(b)(l) of
subpart F of the proposed rule states that "performance
testing, or another form of compliance demonstration," be
performed within 120 days of the compliance date, affective
date, or initial startup, which ever is applicable.  This also
indicates an initial test.
     In reviewing these provisions for response to this
comment, the EPA has determined that the provisions are clear
for performance tests but could be misinterpreted for
compliance determinations for flares.   Section 63.103(b)  was
changed in the final provisions to include initial compliance
determinations and now reads:
     Initial performance tests and initial compliance
     determinations shall be required only as specified
     in subparts G and H of this part.
In addition §63.103(b)(1) of subpart F was changed to include
compliance determinations and now reads:
     Performance tests and compliance determinations
     shall be conducted according to the  schedule and
                             4-17

-------
     procedures in §63.7(a) of the General Provisions and
     the applicable sections of subpart G and H.
     The EPA agrees with the commenters that §63.128(b)(1)(ii)
from the proposed HON is unnecessary and this paragraph has
been removed in the final rule.
     The EPA understands the comment regarding the visible
emissions test to mean that it seems intuitive that visible
emissions will not be present when the flare is shut down.
However, the compliance determination is intended to show that
the facility is capable of operating in conformance with the
requirements specified in §63.11(b) of the General Provisions
which requires that there be less than 5 minutes of observed
visible emissions in any 2 consecutive hours (as well as other
requirements not related to visible emissions).  To do this
there must be visible emissions documentation covering at
least 2 hours.
     Flares occasionally exhibit visible emissions.  Depending
on the type of flare and how it is operated, the most likely
period of visible emissions is during the start-up of a flare
or when there is a rapid increase in the flow of gases to the
flare such as when a second vessel loading is started.  The
compliance determination is usually scheduled in advance and
should be run under representative loading conditions for the
facility; e.g., if multiple vessels are loaded simultaneously
at the facility, the compliance determination should reflect
that the flare is capable of meeting the requirements of
§63.11(b) of the General Provisions, including visible
emissions, under those conditions.  If the routine loading
cycle is less than 2 hours, then an observation period for
that run including the entire loading cycle is acceptable; if
additional loading cycles are initiated within the 2-hour
period, then visible emission observations should be
conducted.  The 2-consecutive-hour period is appropriate for
an inspector observing a suspected violation where it would be
impractical to precisely coordinate loading cycle start and
end.
                             4-18

-------
4.2.2  Monitoring
     Comment:  Two commenters  (A-90-19:  IV-D-70; IV-D-99)
asserted that the HON does not clearly indicate whether or not
all connections between the vapor balancing system and tank
trucks and tank cars will be monitored.  The commenters
(A-90-19:  IV-D-70; IV-D-99) contended that these connections
are likely to be a significant source of toxic fugitive
emissions during loading operations, so they should be
monitored.
     Response;  The transfer operations provisions require
that all connections in a vapor balancing system used to
comply with the rule be inspected.  Since Group 1 racks with
chemicals being vapor balanced may not be considered subject
to the HON, these vapor balancing systems may not be subject
to the inspection requirement under the HON transfer
operations provisions.  Group 2 transfer racks using vapor
balancing for credit under emissions averaging are subject to
the inspection requirements.
     Comment:  Several commenters (A-90-19:  IV-D-58; IV-D-62;
IV-D-73) recommended that the 15-minute monitoring of loading
cycles be made standard for all cycles in the final rule.  One
commenter (A-90-19:  IV-D-58) objected to the requirement for
5-minute monitoring of loading cycles of less than 3 hours
duration.  The commenter (A-90-19:  IV-D-58) claimed that the
proposed HON contains no rationale for more burdensome
monitoring requirements on shorter term loading cycles.  One
commenter (A-90-19:  IV-D-73) contended that the EPA should
set a single frequency rather than tailor it to the loading
cycle or control device operations.   However, the commenter
(A-90-19:  IV-D-73) continued that if this were not possible,
the EPA should retain the options in §63.130(a)(ii).
     Response:  The EPA has reconsidered the monitoring
frequency as requested by the commenters and determined that a
5-minute monitoring frequency is not necessary.  The rule also
requires that daily averages be recorded and reported.
Monitoring at a frequency of 5 minutes instead of 15 minutes
does not significantly increase the assurance that the daily
                             4-19

-------
averages will identify performance that is out of the
parameter ranges.  In fact, allowing a frequency of monitoring
at 15 minutes instead of 5 minutes could result in situations
where the daily average indicates an out-of-range exceedance
when 5-minute monitoring data would not.  This is because •
there may be cases where the owner or operator uses a control
device for a short time during the day; for example, for a
1-hour loading period.  In this case, performing 15-minute
monitoring, the owner or operator will have four data points
to base the daily average on, while 5-minute monitoring would
give 12 data points; if, during this hour, the control device
ran for about 5 minutes outside a monitored parameter range,
one of the four points might indicate the parameter to be out
of the range.  The resulting daily average might be out of the
parameter range.  In the case of 5-minute monitoring, eleven
other data points will most likely be enough to bring the
average into the parameter range.
     It is less likely that the converse would happen, where
the fewer data points under 15-minute monitoring, would
indicate that the daily average is not outside the parameter
range while the 5-minute monitoring would indicate the daily
average is outside the parameter range.  In order for the
daily average to be outside the parameter range, the monitored
parameter would have to be either far out of range or be out
of range for over half the time.  Fifteen-minute monitoring
would be sensitive to either of these situations and is a
sufficient frequency for monitored parameters.  This change
will reduce the monitoring and recordkeeping burden while
still assuring compliance.
     Comment;  One commenter (A-90-22:  IV-D-13) requested the
EPA clarify in §63.126(h) that PRV's needed for safety
purposes may open during loading.  The commenter (A-90-22:
IV-D-13) contended that if a dangerous overpressure situation
develops, the HON should not prohibit PRV's from opening to
prevent an uncontrolled release or even an explosion.
     Response;  The EPA considers the requirement in
§63.126(h) that pressure-vacuum vents not open during loading
                             4-20

-------
to be appropriate.  During loading the vapor collection system
should be collecting vapors with no significant increase in
pressure in the vapor collection system, or in the tank truck
or rail car.  An opening of a safety relief valve indicates
that there is a build up of pressure in the line and
therefore, the vapor collection system is not working
adequately; this can not be allowed.
     As in all operations at a source, emergency situations
can arise requiring emission releases to avoid dangerous
accidents.  The General Provisions provide for a start-up,
shutdown, and malfunction plan in §63.6(e)(3).  This plan is
to include a detailed step-by-step procedure for operating and
maintaining the source during periods of malfunction and a
program for corrective action for malfunctioning process and
air pollution control equipment.  Owners or operators may be
able to include releases from pressure relief valves in their
start-up, shutdown, and malfunction plan if they consider
these releases possible.  However, the releases would have to
be considered a malfunction based on the definition as defined
in the final General Provisions.  The definition in the
proposed General Provisions is as follows:
     Malfunction means any sudden, infrequent, and not
     reasonably preventable failure of air pollution
     control equipment, process equipment, or a process
     to operate in a normal or usual manner.  Failures
     that are caused in part by poor maintenance or
     careless operation are not malfunctions.
4.2.3  Inspections
     Comment:  Several commenters (A-90-19:  IV-D-32; IV-D-33;
IV-D-112) supported the use of DOT certification as a
demonstration of vapor-tightness.  Two commenters (A-90-19:
IV-D-32;.IV-D-112) stated that the HON correctly recognizes
the role of vessel owners in ensuring their vessels are
adequately pressure-tested.  Two commenters (A-90-19:
IV-D-58; IV-D-62)  asserted that a rack owner's or operator's
liability should not extend beyond his area of responsibility,
and recommended that beyond recording properly documented DOT
certification,  the rack owner has no further liability in
                             4-21

-------
connection with the leak.performance of a tank truck or
railcar.  Two commenters (A-90-19:  IV-D-62; IV-D-64)
recommended that EPA clearly state that fulfilling the
recording and recordkeeping requirements associated with the
DOT certification will relieve the owner or operator of the
loading rack from liability for infractions by the
transporter.
     Response:  The transfer operation provisions clearly
state in §63.126(e) that the source owner or operator is
responsible for loading organic HAP's into only tank trucks
and railcars which either have a current DOT certification, or
have been demonstrated to be vapor-tight.  In §63.130(e), the
provisions clearly state the recordkeeping responsibilities of
the source.  The EPA does not consider it necessary or
advantageous to explicitly state that the source has no
further responsibility.  The responsibilities under the
transfer provisions are explicitly stated in the above
mentioned sections of the HON; therefore it is not necessary
to make a broad statement of the sources responsibility.  Also
it is possible that the source has responsibilities under
other rules or in specific circumstances.  Therefore making a
broad statement of the sources'  responsibility could be
confusing; it is not advantageous for the EPA to make a broad
statement in the HON rule regarding the sources'
responsibility for leaks that occur.
     Comment:  One commenter (A-90-19:  IV-D-34 and IV-G-4)
contended that the pressure change used for the vapor-
tightness testing for gasoline trucks is too low compared to
the working pressure of tank cars and tank trucks used by the
chemical industry, and is therefore inappropriate in the HON
rule for transfer operations.  The commenter (A-90-19:
IV-D-34 and IV-G-4) also asserted that EPA-specified gauges
associated with the proposed testing would not be amenable to
the high working pressures and may even be damaged during
testing.
     One commenter (A-90-19:  IV-D-34) suggested that the EPA
develop vapor-tightness testing procedures based on the
                             4-22

-------
maximum allowable working pressures of the tank trucks and
railcars normally used by the chemical industry.  The
commenter  (A-90-19.:  IV-D-34) elaborated on this suggestion in
a subsequent notice  (A-90-19:  IV-G-4) by specifically
suggesting the following:
     "Gas pressure tank trucks to 10 psia and tank cars to
     25 psia.  Bubble test the vessels.  Tighten/repair
     fittings and seals until bubbles are no longer visually
     or audibly detectable."
     Response;  The EPA considers the vapor-tightness test to
be adequate for most situations.  If a facility feels that the
vapor-tightness test is not appropriate for the conditions
that their vehicles operate under, the facility can choose to
load only into DOT-certified tanks.  The facility may also use
another test method if they validate it using Method 301.
These options provide facilities with enough flexibility to
accommodate their specific situation.
     Comment:  Two commenters (A-90-19:  IV-D-70; IV-D-99)
indicated that significant fugitive loading emissions may
result after maintenance activity on transport vehicles if
they are not leak tested.  The commenters (A-90-19:  IV-D-70;
IV-D-99) recommended that the regulation be changed to clearly
require each and every vehicle loaded with a HAP to be
inspected for vapor tightness after each maintenance cycle or
on a semi-annual basis at a minimum, and that visual
inspections of each vehicle and connection be conducted prior
to and during loading.
     Response;  The transfer operations provisions require
annual leak tests.  The EPA considers this frequency adequate
for inspection of tank trucks and railcars for leaks.  The
gasoline marketing study (Evaluation of Air Pollution
Regulatory Strategies for Gasoline Marketing Industry, EPA-
450/3-^84-012a, July 1984, p. 3-11) documents a significant
decrease in leaks from tank trucks that received an annual
inspection.  The commenter did not provide data specifying why
semi-annual inspections were necessary, nor did they describe
their benefits over annual inspections.
                             4-23

-------
     In regards to visual inspections of each vehicle and
connection prior to and during loading, operators are required
to operate the vapor collection system in order to collect the
organic HAP vapors displaced during loading §63.126(a)(1) and
therefore must connect the vapor recovery hoses properly to
the vehicles.  It is not necessary to specifically require
that this be inspected.  If the commenter is concerned that
leaks will occur during loading due to a failure in the
equipment, these will be difficult to visually identify.
Also, during loading, the vapors will be collected through the
vapor recovery line and most of the vapors will take this path
of least resistance instead of through any failure in the
equipment.  The EPA considers the annual vapor tightness test
or DOT certification to be sufficient.
     Comment:  One vendor (A-90-19:  IV-D-8) provided
information to the EPA on a leak detection device they claimed
could be used instead of Method 21 for compliance with the
inspection provisions for collection systems.  Others
(A-90-19:  IV-D-14; IV-D-15; IV-D-17; IV-D-18; IV-D-19;
IV-D-20; IV-D-23; IV-D-24; IV-D-25; IV-D-27; IV-D-28),
(A-90-20:  IV-D-2; IV-D-4) commented on procedures and
requirements of Method 21.
     Response:  The EPA has provided a discussion on Method 21
and the alternate leak detection device in section 5.0 of this
BID volume.
     Comment:  Several commenters  (A-90-19:  IV-D-32; IV-D-33;
IV-D-73),  (A-90-22:  IV-D-7) suggested that the requirements
in subpart H for leak detection of equipment are applicable to
vapor collection systems or transfer racks.  The commenters
(A-90-19:  IV-D-32; IV-D-33; IV-D-73), (A-90-22:  IV-D-7)
recommended deleting the requirement in §63.126(a)(3)(i) and
§63.126(b)(3)(ii) of the transfer provisions.  One commenter
(A-90-19:  IV-D-73) maintained that in many SOCMI plants there
is a plant-wide closed-vent collection system and that it is
essential that there be only one regulatory requirement for
that system.
                              4-24

-------
     One commenter  (A-90-19:  IV-D-33) anticipated  that  in  the
future the subpart H requirements may also apply to non-HON
processes.  The commenter  indicated that  it would be most
expedient and less burdensome if there were only one
requirement for leak detection for all MACT standards.
     One commenter  (A-90-19:  IV-D-73) contended that the HON
transfer repair provisions would not allow for delay of  repair
which is needed for components in HAP service or in closed-
vent service.  Two commenters (A-90-19:   IV-D-32; IV-D-73)
requested that the EPA add a provision for delay of repair
because transfer vapor collection systems may be common  with
other processes.
     Response:  Method 21 leak inspection provisions are
assembled into a new section, §63.148 of  subpart G.  The
transfer operations provisions have been  revised to refer to
the closed-vent system provisions in this new section,
§63.148.
     The transfer operations provisions have also been revised
to refer to the delay of repair provisions in §63.148 of
subpart G.
     Comment:  Three commenters (A-90-19:  IV-D-64;  IV-D-73)
(A-90-22:  IV-D-13) asserted that the EPA should exempt
equipment from transfer provisions consistently in  all
sections.  Two commenters  (A-90-19:  IV-D-73) (A-90-22:
IV-D-13) contended that the intent of §63.126(i) is  to exclude
PRV's for safety purposes from the requirements to  have  a flow
indicator or car seal.  The commenters (A-90-19:  IV-D-73)
(A-90-22:  IV-D-13) stated that bleeds, drains, etc. are
excluded from the car seal or lock requirements by  §63.127(d);
however, no such exclusions are in §63.126(i).  The  commenters
(A-90-19:  IV-D-73) (A-90-22:  IV-D-13) urged that these two
sections be aligned.  Another commenter (A-90-19:   IV-D-73)
requested that the EPA carry over the exclusion in  §63.127(d)
to the reporting and recordkeeping section [§63.129(d)].
     Response;  Sections 63.126(i), 63.127(d), and  63.129(d)
have been revised to exclude the same equipment.
                             4-25

-------
4.2.4  Compliance Schedule
     Comment;   Several commenters (A-90-19:  IV-D-32; IV-D-34;
IV-D-58; iv-D-62; IV-D-112) claimed that it may not be
possible to test leaks within the 15-day period directed in
the HON because of the intermittent nature of transfer
operations.  The commenters (A-90-19:  IV-D-32; IV-D-34;
IV-D-58; IV-D-62; IV-D-64; IV-D-112) protested that if a leak
has been worked on but the transfer operation component is
shut down because all loading has been completed or if the
rack is operated sporadically, there would be no way to screen
the component to determine if it had been fixed until the next
transfer operation, which may exceed 15 days.
     Several commenters (A-90-19:  IV-D-32; IV-D-34; IV-D-58;
IV-D-62; IV-D-112) suggested that the EPA should allow
monitoring to take place within the 15 day limit after repair,
or at the beginning of the next transfer operation if loading
operations have ceased before screening could take place.
Three commenters  (A-90-19:  IV-D-32; IV-D-34; IV-D-62) claimed
that since the transfer rack would not be in operation, there
would be little to no escape of organic vapors to the
atmosphere during idle time.  One commenter (A-90-19:
IV-D-63) requested that the EPA grant an automatic extension
or exemption from repairing leaks within 15 days of detection
for low quantity transfer operations that may only be used
once or twice a month.
     Response;  The transfer operations provisions have been
revised to allow for testing leaks within the 15-day limit
after repair,  or at the beginning of the next transfer
operation if loading operations have ceased by the time
screening would have taken place.
4.3  RECORDKEEPING AND REPORTING
     Comment;   Two commenters (A-90-19:  IV-D-58; IV-D-64)
contended that the recordkeeping and reporting requirements in
the HON transfer provisions were excessive and burdensome.
One commenter (A-90-19:  IV-D-92) specifically asserted that
the transfer provisions require excessive recordkeeping if the
facility does not have a flare.
                             4-26

-------
     One commenter  (A-90-19:  IV-D-34) contended that
recordkeeping and reporting requirements are excessive  for
Group 2 transfer operations and puts the source in the
position of being required to keep extensive records and make
calculations on a point for which controls are not applicable.
The commenter (A-90-19:  IV-D-34) asserted that recordkeeping
is a needless requirement for Group 2 racks since the
determination of applicability for individual racks is  to be
made based on utilization that occurred during the year
preceding promulgation [§63.100(b)(5)(v) in the proposed HON].
The commenter (A-90-19:  IV-D-34) claimed that the
applicability does not change unless there is a change  in the
material loaded at the loading arm or hose, which requires the
applicability to be redetermined.  The commenter (A-90-19:
IV-D-34) expressed concern that extensive recordkeeping and
reporting requirements subject the source to potential
noncompliance for failure to maintain records for a source
which needs no control.
     The commenter  (A-90-19:  IV-D-34) suggested that the EPA
delete recordkeeping requirements for Group 2 racks and
instead require a certification in a semi-annual report that a
change in material loaded has not occurred.
     Response:  The EPA has made every effort to reduce the
recordkeeping and reporting burden and to require only  those
records and reports necessary to determine compliance.  For
example, the 5-minute monitoring and recordkeeping frequency
for control devices has been changed to a 15-minute monitoring
and recordkeeping frequency.  The recordkeeping and reporting
requirements for all the various control devices reflect what
is necessary to determine compliance.
     The EPA does not consider the recordkeeping and reporting
requirements excessive for Group 2 transfer racks.   The
records are very limited; only design and actual throughput,
weight-percent organic HAP,  and the rack-weighted partial
pressure of chemicals transferred are required.   These  records
are necessary to determine if there has been a change in the
rack's group status.  The commenter is incorrect in stating
                             4-27

-------
that group status is determined based on pre-promulgation
data.  If the amount and/or specific chemicals loaded changes
after promulgation, a Group 2 rack could become Group 1.
     The commenter has misinterpreted §63.100(b)(5)(v) in the
proposed HON and §63.100(h) in the final rule.  The intent of
proposed §63.100(b)(5) [§63.100(h) in the final rule] is to
explain how to assign a rack/arm which transfers chemicals
from more than one chemical manufacturing process.  Once the
rack/arm is assigned to a chemical manufacturing process, it
can be determined if the chemical manufacturing process, and
therefore the rack/arm is subject to subparts F and G.  Once
it is determined that a rack/arm is subject to subpart G, then
group status is determined.  Therefore, if an owner or
operator determines that a rack/arm is not subject to
subparts F and G based on proposed §63.100(b)(5)(v)
[§63.100(h) in the final HON], then the rack/arm is not
subject to the requirements of either Group 1 or Group 2
racks.  This section of subpart F has been revised to make
this more clear.
     Comment:  One commenter  (A-90-19:  IV-D-34) contended
that recordkeeping requirements for continuous monitoring of
transfer operations are excessive and costly.  The commenter
(A-90-19:  IV-D-34) claimed that loading racks are physically
remote from control rooms in some facilities and there is no
location where data can be easily stored.  As a result,
intrinsically safe, weatherproof recorders would be required.
The commenter (A-90-19:  IV-D-34) claimed that simpler control
devices would operate well with a reduced level of monitoring
that may allow a less expensive and more practical way of
ensuring proper emissions control.
     The commenter (A-90-19:  IV-D-34) suggested that the
frequency of monitoring be significantly reduced to the extent
that it may be done without expensive continuous data
collection systems.
     Response:  The monitoring frequency has been reduced from
5 minutes in some cases to 15 minutes in all cases.  The EPA
recognizes that some new equipment may be required to meet the
                              4-28

-------
monitoring, recordkeeping, and reporting requirements of the
transfer operations.  However, this frequency is considered
necessary for ensuring compliance.
     The commenter did not provide any specific
recommendations for reducing the burden or cost, except for
the suggestion to reduce frequency.  As explained in the
recordkeeping. and reporting chapter of the BID, an owner or
operator may request site-specific approval to use non-
automated monitoring systems if relevant operating parameters
are read and recorded no less frequently than once per hour,
and daily average values are calculated from the hourly values
and recorded, as provided in §63.151.  The request must
contain:  (1) a description of the planned monitoring and
recordkeeping system; (2) documentation that the source does
not have an automated system; (3) reasons the source is
regulating an alternative monitoring and recordkeeping system;
and (4) demonstration that the proposed monitoring frequency
is sufficient to represent control device operating conditions
considering typical variability of the specific process and
control device operating parameter being monitored.  In
approving the request, the permit authority may consider the
variability of the parameter, and whether a monitoring
frequency that is longer than once every 15 minutes is
sufficient to characterize control device operation.
4.4  WORDING OF THE PROVISIONS
     Comment:  One commenter (A-90-19:  IV-D-64) contended
that throughout the transfer provisions,  the word "recorder"
should be changed to "continuous recorder", since "recorder"
is an undefined term and "continuous recorder" is defined.
     Response:   At proposal,  the term "continuous recorder"
was specifically avoided because it refers to a device capable
of generating a record at least once every 15 minutes.   In the
proposed transfer operations provisions,  some records are
taken every 5 minutes.  Since this was changed in the final
transfer operations provisions to records at least once every
15 minutes,  all references to "recorder"  have been changed to
"continuous recorder."  As with the other emission points, if
                             4-29

-------
there are no monitoring parameter excursions during the day,
the owner or operator has the option to retain hourly averages
and discard the 15-minute records.
     Comment;   One commenter requested that the EPA provide
English unit equivalents wherever a metric unit appears.
     Response;  The regulation specifies only metric units
because the EPA enforces standards based on the metric system.
Conversions to English units would introduce imprecision and
lead to situations where enforcement is unclear.
4.5  MISCELLANEOUS
     Comment;   Two commenters (A-90-19:  IV-D-70;  IV-D-99)
asserted that the HON should make provisions for the operator
to interlock the- flow indicator with an automatic system to
stop the loading procedure and close all open lines when the
flow meter indicates a leak to the atmosphere.
     Response;  The flow indicator, car seal or lock-and-key
closures required in §63.126(i)  ensure that emissions are not
diverted to the atmosphere, directly or indirectly, through a
valve in the vent system.  The EPA does not consider it
necessary to require an automatic system with an interlocking
flow indicator; however this type of system is not precluded
in the transfer provisions.
     Comment;   One commenter (A-90-19:  IV-D-75) asserted that
the point of generation for loading operations should be
defined as after the point where the stream is destined for
disposal, because that point is where emissions may occur.
The commenter (A-90-19:  IV-D-75) contends that this is the
only reasonable definition, and should be adopted.
     Response:  The EPA does not consider a definition for
point of generation to be applicable or necessary to transfer
operations.  The commenter did not give any details on why
point of generation should be defined for transfer operations.
     Comment;   Two commenters (A-90-19:  IV-D-61; IV-D-92)
recommended that the HON include more general language to
allow for flexibility in controlling vapors from vapor
balancing.  One commenter  (A-90-19:  IV-D-92) claimed that the
vapors from vapor balancing may be routed back to the process
                             4-30

-------
unit, or may be pressurized to the tank car.  Another
commenter  (A-90-19:  IV-D-61) contended that the HON should
allow vapor collection and return to process units as an
alternative control technology for transfer emissions.  The
commenter  (A-90-19:  IV-D-61) stated that, for vapor balancing
systems that return material to the process unit, the vapor
collection system subject to LDAR requirements [such as
specified in proposed §63.126(a)(3)] should be defined as the
vapor handling equipment up to the point of commingling with
raw feed.
     Response;  The transfer operations provisions allow for
owners or operators to combine vapors with process vent
streams which are then sent to a control device.  There is
nothing in the provisions that precludes shared control
devices.  Also, transfer operations that are under pressure
are not subject to the HON transfer provisions.
     The EPA considers allowing vapors to be recycled back to
the process unit to be acceptable, except in cases where the
vapors are only being vented through a process unit and out to
the atmosphere.  In order to allow for recycling back to the
process unit, an option has been added to the transfer
operations provisions allowing vapors from transfer operations
to be commingled with the raw feed.
     Comment;  One commenter (A-90-19:  IV-D-64)  asserted that
vapor collection systems should be operated to minimize,
rather than prevent, the incidence of organic HAP vapors
collected at one loading arm from being passed through another
arm, because it would be impossible to completely prevent
minor occurrences of this.
     Response:  The vast majority of vapors can be prevented
from passing from one arm through another to the atmosphere;
however the EPA considers requiring all HAP vapors from
loading arms be prevented from being lost to the atmosphere
overly stringent.  The EPA agrees with the commenter that it
is impossible to completely prevent every molecule of HAP
vapors from.being diverted through another arm to the
                             4-31

-------
atmosphere.  In order to.better communicate this requirement
the provisions in §63.126(a)(2) have been revised.
     Comment;   One commenter (A-90-19:  IV-D-33) suggested
that the definition of recovery device in §63.101 be used in
§63.111.
     Response:  Recovery device is defined in much the same
way in §63.101 of subpart F and §63.111 of subpart G.  In
order to eliminate redundancy,  the definition in §63.111 of
subpart G was removed in the final provisions.
     In order to be consistent, the definition for control
device was moved from §63.111 of subpart G to §63.101 of
subpart F.
     Comment;   One commenter (A-90-22:  IV-D-13) suggested
that the EPA clarify that manifolded vent lines on loading
arms do not require car seals or flow indicators.  The
commenter  (A-90-22:  IV-D-13)  stated that, for manifold vent
lines, a positive closure such as a plug or a cap is preferred
over car seals because car seals are impractical.  The
commenter  (A-90-22:  IV-D-13)  explained that car seals might
need to be removed and reapplied several times at a busy
loading rack.   The commenter (A-90-22:  IV-D-13) contended
that monthly inspections of car seals and flow indicators on a
manifolded vent line are meaningless.
     Response:  The provisions requiring a car seal, lock-and-
key type closure, or a flow indicator do not apply to
manifolded vent lines on loading arms.  The provisions apply
to the vapor recovery lines.
                             4-32

-------
                      5.0   EQUIPMENT LEAKS

5.1  STANDARDS
5.1.1  363.162;  General
     Comment;  Several commenters  (A-90-19:  IV-D-33; IV-D-34;
IV-D-57; IV-D-73; IV-D-77; IV-D-79; IV-D-97) requested that a
sufficient period of time should be allowed for installation
of equipment required to achieve compliance with the standard.
One commenter  (A-90-19:  IV-D-73) argued that the 6 to
18 months allowed in the proposed rule did not take into
account implementation problems that might occur.  The group
of commenters  (A-90-19:  IV-D-33; IV-D-34; IV-D-57; IV-D-73;
IV-D-77; IV-D-79; IV-D-97) requested a compliance time similar
to the 3-year compliance schedule allowed under subpart G.
Another commenter (A-90-19:  IV-D-33)  argued that subpart H
should allow up to one-year compliance waiver for installation
of controls.  Examples of equipment requirements cited include
installation of a seal system on a compressor, installation of
a sampling connection system or equipping a open-ended valve
or line with a cap.
     Response;  The EPA does not agree with the commenters
that sources should be allowed up to 3 years to comply with
the provisions in subpart H.  Subpart H consists of a
combination of work practice requirements for many equipment
components and equipment standards for compressors, sampling
systems, open ended lines or valves, and pressure relief
valves.  Unlike the requirements in subpart G, the equipment
required by subpart H should not involve long periods of time
for design, construction, and installation.  The commenters
did not provide any information that would justify
establishing a source-category-wide compliance schedule for
                              5-1

-------
subpart H similar to that provided for subpart G.  The EPA
recognizes that there may be circumstances present in
individual facilities where an extension is appropriate for
compliance with certain requirements in subpart H.  In such
cases, the owner or operator may request an extension of
compliance through the provisions of §63.6(1)(4) of subpart A.
Section 63.182(a)(6) of subpart H has been added to subpart H
to clarify that extensions of compliance may be requested if
additional time is necessary for installation of equipment
required by subpart H.
     Comment:  One commenter (A-90-20:  IV-D-19) suggested the
general standards for process units subject to subpart H were
ambiguous and needed clarification.  Specific concerns
mentioned by the commenter (A-90-20: IV-D-19) were (1)
paragraph (a) requires the owner or operator to "demonstrate
compliance" but actually is requiring compliance, not the
demonstration thereof; (2) paragraph  (b) needs to be more
explicit as to what records could be reviewed; and (3)
paragraph (e) needs to be clarified to specify that it refers
to equipment intended to operate under vacuum during normal
operations, not that the equipment must be under vacuum at all
times including startup or shutdown.  In addition, paragraphs
(e) and (f) require negative recordkeeping - identification of
individual pieces of equipment that are not subject to the
subpart.
     Response:  The EPA considered the commenter's suggestions
and, where appropriate, revised the final rule.  Specific
changes made to the general standards in §63.162 were:   (1)
Paragraph (a) was deleted because all general compliance
requirements are located in subparts F and I;  (2) paragraph
(b) revised to cite the specific records; and  (3) paragraph
(e) was deleted.  Paragraph (b).was revised to specifically
cite the required records to remove any possible ambiguity in
the rule regar  _ng required records.  Section 63.181  in the
final rule specifies all the required records.  The
requirement in paragraph (e) to document all equipment in
vacuum service was removed because it is possible to  identify
                              5-2

-------
such equipment by  inspection of the process unit and this
requirement added  an unnecessary recordkeeping burden.  The
requirement in paragraph  (f) of the proposed rule to document
equipment in HAP service  less than 300 hours was retained,
however.  It was the EPA's opinion that this record was
necessary because  it would not be possible to determine this
by  inspection.
     Comment:  One commenter (A-90-19:  IV-D-73) asserted that
it  is arbitrary to require l-year compliance waiver requests
to  be submitted 1  year before the compliance date.  The
commenter (A-90-19:  IV-D-73) requested that the EPA
reconsider these stipulations.
     Response;  The EPA believes that the commenter
misinterpreted the requirement in §63.182(a)(2) for submittal
of  a notification  of applicability as also applying to waiver
requests.  Section 63.182 has been redrafted to remove this
source of confusion and specify the dates by which the request
must be submitted  and the required information.  The final
rule specifies that the waiver request must be submitted no
later than 3 months before the compliance date.  This time
period is sufficient to permit review of the application and
notification before the compliance dates.  The 'submittal date
differs from the time period specified in the general
provisions (subpart A) because the compliance dates for
subpart H and the  provisions in subpart A would not allow any
requests for compliance extensions.
     Comment:   Two commenters (A-90-19:  IV-D-79; IV-D-105)
reasoned that facilities subject to 40 CFR part 61 subpart F
should be exempt from 40 CFR part 63 subpart H.  The
commenters (A-90-19:  IV-D-79; IV-D-105) argued that no
benefit would be gained by making these facilities comply with
subpart H since they are already subject to a similar program.
One commenter (A-90-19:  IV-D-79) also requested that the EPA
provide a phase-in period for facilities currently complying
with existing equipment leak rules.
     Response;  As part of the general evaluation of
overlapping requirements in part 60 and 61 rules, the EPA
                              5-3

-------
considered whether the equipment leak standard in the Vinyl
Chloride NESHAP (subpart F of part 61)  was more stringent than
the requirements in subpart H.  It was concluded that for
Vinyl Chloride the stringency comparison needs to be on a
case-by-case basis.  It is not possible to do the evaluation
on a national basis because subpart F of part 61 initially
required a self-developed program and the requirement to
comply with the provisions in subpart V of part 61 was added
later, as an additional requirement.  Therefore, the final
rule provides that sources subject to both subpart F of
part 61 and the HON may request a determination by the permit
authority of the program to be implemented.  Because the phase
I provisions for pumps and valves are identical to those in
subpart V of part 61, the EPA believes that this evaluation
can be conducted during the first year of the standard and no
additional burden will result.
     The EPA did not provide a phase-in period for facilities
currently complying with existing equipment leak standards in
part 60 or 61, as requested by the commenter.  The EPA thinks
a facility that is already complying with existing rules
(i.e., NSPS or NESHAP) should have less difficulty achieving
compliance under subpart H because they already have a program
in place.  The commenter did not provide details on why extra
time should be allowed for these facilities.
     Comment:  One commenter  (A-90-20:  IV-D-19) suggested
that it is not necessary to exclude dual mechanical seal pumps
and compressors in VOC service from the override of 40 CFR
part 60 or 61 requirements in §63.160(d).  The commenter
observed that requirements for dual mechanical seal pumps and
the requirements for compressors are virtually identical and
thus, there is no need to require such exceptions.
     Response;  The standard has been revised to remove this
limitation because, as noted by the commenter, there are no
practical differences in the requirements.
     Comment;  Two commenters (A-90-20: IV-D-38; IV-G-3)
argued that application of the requirements of subpart H to
phosgene-containing equipment at diisocyanate units would not
                              5-4

-------
be useful since these units are presently monitored for any
leaks using sensitive perimeter monitoring systems.  The
commenters suggested that subpart H include a provision that
would allow use of area monitoring systems for phosgene
containing equipment in lieu of the leak detection provisions
that would otherwise apply under subpart H.
     Response;  The final rule provides provisions that allow
establishment of alternative monitoring provisions provided it
can be demonstrated that this system can at least detect a
500 ppm leak.  This demonstration can be based on dispersion
modeling, engineering calculations, or past experience.  It is
expected that allowable systems will be highly dependent on
the HAP's being monitored as well as site layout.
     Comment;  One commenter (A-90-20:  IV-D-27) recommended
that an exemption from periodic monitoring be provided for
cases where it is not feasible to monitor.  The commenter
cited 2 examples of cases where monitoring would be
infeasible:   (1) chemicals which can not be reliably detected
by available instruments; and (2) containment areas where the
process is isolated due to concerns with health and safety
issues or concerns with product contamination.  The commenter
suggested for these cases the rule require repair if there is
visible, audible or olfactory evidence of a leak.
     Response;  The EPA does not believe that exemption from
periodic monitoring is warranted or necessary.  For cases
where no instrument exists, the rule allows the owner or
operator to monitor a surrogate or to request approval of an
alternative program.  Section 63.179 of subpart H also exempts
enclosed process units that are vented through a control
device from the periodic monitoring requirements.
     Comment;  One commenter (A-90-19:  IV-D-86) stressed that
6 months to implement the rule for Group I sources is too
brief.  The commenter (A-90-19:  IV-D-86) recommended that
small facilities be allowed 3 years to comply, citing the lack
of environmental staffing and the cost of instituting the
program as reasons to defer implementation.  Another commenter
(A-90-19:  IV-D-92)  also noted that a 6-month period was too
                              5-5

-------
brief to implement the rule for facilities that are not yet
implementing an LDAR program.
     Response;  The EPA considers that ample notice has
already been given.  The original agreement was published in
the Federal Register on March 2, 1991.  The HON was proposed
in December 1992 and will be promulgated at the end of
February 1994.  This is a time span of 3 years and the EPA
maintains that this should have provided ample time to
determine applicability and implement the means for achieving
compliance.
     Comment;   One commenter (A-90-20:  IV-D-27) recommended
that the definition of connector be modified to state that
connections between sections of a vessel and between the
vessel and head gaskets are not considered connectors.  This
commenter also suggested that the definition of screwed
connector be modified to incorporate the definition of
connector.  The commenter thought that this change would avoid
confusion.
     Response;   The EPA considers the two definitions to be
clear and that the possibility of confusion unlikely.  The
definition of connector states that:  "connector means
flanged, screwed, or other joined fittings used to connect two
pipe lines or a pipe line and a piece of equipment. . "  In
other words, a connector is a device that connects two pipes
or a piece of equipment.  Since a vessel is neither a piece of
pipe or a piece of equipment, connections between sections of
a vessel and between the vessel and head gaskets cannot be
considered connectors.  The rule defines equipment as:
     Equipment means each pump, compressor, agitator,
     pressure relief device, sampling connection system,
     open-ended valve or line, valve, connector, surge
     control vessel, bottoms receiver, and
     instrumentation system in volatile hazardous air
     pollutant service; and any control devices or
     systems required by this subpart.
The definition of connector was not added to the definition of
screwed connector because the definition of connector clearly
states that screwed connectors are one type of connector.
                              5-6

-------
     Comment;  One commenter  (A-90-20:  IV-D-19) suggested
that all units, not just batch processes, be allowed to
monitor anytime the equipment is in service with any
detectable material, not just VHAP.  The commenter suggested
this additional flexibility would make the monitoring more
cost-efficient.
     Response:  The EPA agrees with the commenter that this
flexibility should be available for continuous as well as
batch processes.  Accordingly the rule has been modified to
allow use of surrogate monitoring to check for leaking
equipment.
5.1.2  S63.163;  Pumps in Light Liquid Service
     Comment;  One commenter  (A-90-19:  IV-D-60) requested
that the rule allow an owner or operator to calculate a
percentage of leaking valves, connectors, or pumps for groups
of similar process units, citing this option as one method of
providing meaningful leak rate data.
     Response;  The EPA is uncertain just what the commenter
meant by "meaningful leak rate data".  The pump standard, in
§63.163(a)(2) of subpart H, allows the owner or operator to
calculate the percent leaking pumps on a process unit basis or
a source-wide basis.  The Committee agreed to this provision
to consider the small number of pumps typically in a process
unit and potential problems associated with small populations
and site-specific concerns.  The commenter's suggestion of
another option for the calculations does not appear necessary
and will add additional complexity to the standard.
     Comment;  One commenter  (A-90-19:  IV-D-89) recommended
that §63.163(g) of subpart H be modified to also exempt from
the monitoring requirements in paragraphs §63.163(b) through
(e)  any systems that capture and transport leakage from the
seal(s)  to the process recovery system.
     Response;  The provisions in §63.163(g) have been revised
to allow the owner or operator to route the leakage back to
the process where the material will be recycled as well as to
a control device.   The EPA considers this change to be a
                              5-7

-------
clarification since the process recovery system meets the
definition of control device under subpart H.
     Comment;   One commenter (A-90-20:  IV-D-19) thought the
rule should specify that pumps installed after the
applicability date are in the same phase as the remainder of
the process unit.  The commenter (A-90-20:  IV-D-19)  suggested
that paragraph §63.163(a)(3) be added to read:
     (3)  All pumps within a process unit are in the same
     phase, including pumps installed in the process unit
     after that applicability date.
     Response;  The EPA does not think that the suggested
language needs to be added to the standard.  Subpart H
specifies that the phases are determined on a process unit
basis not by the individual piece of equipment.  The EPA
believes that the commenter's concern result from difficulties
with implementation of the provisions in 40 CFR part 61,
subparts J and V, benzene equipment leaks NESHAP.  The benzene
NESHAP specifies applicability in terms of each piece of
equipment and not on a process unit basis.  To prevent similar
difficulties from arising in implementation of subpart H, the
EPA will explain in enabling materials and inspection manuals
that the phases of the pump and valve standards are determined
on a process unit basis and addition of a new valve or pump
does not alter the phase the equipment is considered to be in.
     Comment;   One commenter (A-90-20:  IV-D-19) requested
clarification of when follow-up monitoring is required for a
pump that has been repaired.  The commenter  (A-90-20:
IV-D-19) suggested amending §63.163(c)(1) by adding the
following:
     Repaired shall mean that indications of liquids
     dripping from the pump seal are no longer present
     when the pump is returned to VHAP service.
     Subsequent monthly monitoring may be used to confirm
     that repair was successful.
The commenter reasoned since monitoring is performed monthly,
monitoring during the next scheduled monitoring period should
be acceptable.
     Response;  The term repaired is defined in §63.161 as
"equipment is adjusted or otherwise altered to eliminate a
                              5-8

-------
leak as defined  in the applicable sections of this subpart."
Thus, a pump cannot be classed as repaired until it is
monitored and  is confirmed to be below the action level.  This
confirmation monitoring  is an inherent part of the LDAR
program and should not present an undue burden.  The EPA does
not consider the suggested change is necessary.
     Comment:  One commenter  (A-90-20:  IV-D-19) recommended
that the EPA not require monitoring of DMS pumps when a leak
is determined  visually.  The commenter (A-90-20:  IV-D-19)
observed that  §63.163(b)(3) specifies that indications of
liquids dripping found during a visual inspection of a pump
are considered leaks.  The commenter (A-90-20:  IV-D-19)
thought that this a sensible approach, which eliminates
unnecessary extra monitoring, but is not consistent with
§63.163(e)(4)(i).  The latter section requires monitoring if a
DMS seal pump  shows indications of liquids dripping at the
time of the visual inspection.  The commenter  (A-90-20:
IV-D-19) recommended that the EPA make the two sections
consistent.
     Response;   The two cited sections in the pump standard
differ because the DMS seal leak could be a loss of barrier
fluid and would  not result in loss of volatile materials or an
instrument reading of 1000 ppm.  In such cases, the DMS seal
would not be considered to be leaking.  The provisions in
§63.163(b) of  subpart H apply pumps such as single mechanical
seal pumps, reciprocating pumps, etc.  In these cases, the
presence of a  drip will indicate loss of process fluid that is
in light liquid  service and undoubtedly would be measured as a
leak.  The EPA,  therefore, did not revise the provisions as
suggested because some owners or operators subject to the
standard would object to the loss of the opportunity to show
that the drip  from a DMS is not a leak.
     Comment;   One commenter  (A-90-20:  IV-D-19) suggested
that the percent leaking pumps calculation is not appropriate
for process units/plant sites with a large number of DMS or
sealless pumps.  The commenter (A-90-20:  IV-D-19) claimed
that some process units/plant sites have a large number of DMS
                              5-9

-------
and sealless pumps.  Since these pumps are not included in the
PL term of the percent leaking pumps calculation (monitoring
required by §63.163(e)(4)(i) is not included in the definition
of PL), a plant with over 90 percent non-single seal pumps
would never exceed the 10 percent limit which triggers the
pump QIP.  Plants in this situation should not be required to
calculate percent leaking pumps, as the data is meaningless.
The commenter  (A-90-20:  IV-D-19) suggested exempting these
plants from the requirement to calculate percent leaking
pumps.
     Response:  The EPA agrees that in such cases the
calculation is unnecessary.  The recommended language was
added to the pump provisions in §63.163 of subpart H.
5.1.3  §63.164;  Compressors
     Comment;  One commenter (A-90-19:  IV-D-92) recommended
that the use of double mechanical seals should not be
required.  Instead, the commenter (A-90-19:  IV-D-92)
advocated that the standard be expressed as a performance
standard that would necessitate use of the appropriate seal.
     Response;  As with pumps, a performance standard for
compressors is not feasible.  Even though compressor seals can
be equipped to release emissions into a conveyance mechanism,
measurement of these emissions is impracticable.  The standard
allows use of systems that vent the seal area to a control
device as well as dual mechanical systems or sealless
compressors.
     Comment:  One commenter (A-90-19:  IV-D-92) asserted that
compressors with double mechanical seals should be exempt from
monitoring requirements.
     Response;   Compressors equipped with double mechanical
seals are exempt from the LOAR program.  The only requirement
is for a sensor to detect failure of the seal system or
barrier fluid system.  These sensors are necessary because
seals can fail and large emissions could result.
     Comment;  One commenter (A-90-20:  IV-D-19) recommended
several clarifications to the provisions for compressors
vented to closed-vent systems.  The commenter  (A-90-20:
                             5-10

-------
IV-D-19) requested clarification of the type of enclosure that
would meet the criteria for exemption:  Would a laboratory
type hood be considered applicable, or does the enclosure need
to fully encase the equipment?  The commenter (A-90-20:
IV-D-19) also requested that the closed-vent system be allowed
to vent back to the process, as well as to a control device.
The commenter (A-90-20:  IV-D-19) cited provisions in the
benzene waste operations NESHAP [See 40 CFR 61.342(c)(1)(iii)
of subpart FF of part 61] as an example where the EPA has
allowed recycling in lieu of destruction only.  The commenter
(A-90-20:  IV-D-19) questioned the relevance of the
requirements of paragraphs 63.164(c) through (f) of subpart H
for compressors equipped with closed vent systems.  It was
also noted that the last phrase of §63.163(h) "except as
provided in paragraph (i) of this section" is unnecessary.
Paragraphs 63.164(h) and (i) are separate exemptions, not
dependent on each other.
     Response;  The provision in §63.164(h) is deliberately
drafted in a manner that does not specify the actual equipment
that can be used.  That is, the standard does not specify that
the compressor be fully enclosed or equipped with a hood, or
that other specific equipment be applied.  The requirement is
to collect any leakage and convey it to a control device.
This requirement can be met by any number of different
systems.  Systems that enclose ports in the seal area and
evacuate the collected gases are one acceptable means of
compliance as are systems that enclose the entire compressor.
While the EPA understands the desire for more specificity as
to what is acceptable means of compliance, the EPA is also
concerned that the standard be achievable by a number of
different systems and allow flexibility.  Therefore, the
language in §63.164 was not revised to be more explicit as
requested by the commenter.
     In response to the commenter's concerns, §63.164(h) was
edited to allow venting of emissions to the process or to a
control device.   The cross references to the other provisions
                              5-11

-------
in §63.163 were also revised to clarify the exempted
paragraphs.
5.1.4  363.165;  Pressure Relief Valves in Gas/Vapor Service
     Comment;  One commenter (A-90-20:  IV-D-19) thought the
monitoring requirements for pressure relief devices in
gas/vapor service to be unclear.  Specific questions raised by
the commenter  (A-90-20:  IV-D-19) were:  Did the EPA intend
for pressure relief devices to be monitored initially in order
to determine that they meet the less than 500 ppm above
background criteria, and if not why does §63.181(b)(4) require
retention of documentation of compliance tests required in
§63.165.  If it was the EPA's intention to require such
monitoring, the commenter (A-90-20:  IV-D-19) recommended that
it be written out in this section.
     Response;  The intent of the provisions in §63.165(a) is
to demonstrate that the PRV has reseated properly after an
overpressure discharge.  The standard does not require a
compliance test or routine monitoring.
     Comment;  One commenter (A-90-20:  IV-D-19) requested
that the EPA clarify whether pressure release events are
isolated incidents or can occur as a series of discharges over
a relatively short time period.  For process units that cannot
be shutdown immediately upon upset, the commenter thought the
present language in §63.165(b)  would require monitoring with
no benefit under potentially dangerous conditions.  Another
commenter (A-90-19:  IV-D-34) requested that §63.165(b)(2) be
edited to require monitoring within 5 days of being repaired
or returned to VHAP service.
     Response;  The EPA believes that the provisions  in
§63.165(b) already accommodate the situation described by the
first commenter since delay of repair is allowed.  The purpose
of the monitoring is to confirm that the PRV has reseated
properly after an overpressure discharge.  The definition of
pressure release has been revised to clarify that  it may be a
single isolated discharge or a series of releases  over a short
time period due to the same process malfunction.
                             5-12

-------
       The EPA agrees with the other conunenter's that the time
 period following a process unit shutdown should be clarified.
 The provisions in §63.165(b)(2) have been revised to clarify
 that the 5 days is following repair and being returned to HAP
 service.
      Comment;   Two commenters (A-90-19:  IV-D-32; IV-D-34)
 suggested that pressure relief valves which are unsafe to
 monitor or repair should be exempted from these requirements.
 One of the commenters (A-90-19:  IV-D-32)  noted that a major
 safety risk would be posed if a pressure relief valve in
 gas/vapor service released while monitoring or repair
 personnel were present.
      Response;  The intent of the provisions in §63.165 is to
 ensure that there is no leakage from the PRV during normal
 operations (i.e.,  periods when there is not an overpressure
 discharge).   There are  two primary alternatives for
 controlling equipment leaks from PRV's:  use of a rupture disk
• in conjunction with the PRV,  or use of a closed vent system.
 If an owner or operator elects to use either of these control
 options, there is no requirement to monitor the PRV after
 repair.
      The standard also  allows use of a PRV alone provided the
 PRV is demonstrated to  have been returned to less than 500 ppm
 within 5 days  of the overpressure discharge, unless repair is
 technically infeasible  without a process unit shutdown.  While
 the commenters did not  provide examples of situations where it
 would be unsafe to repair or monitor the PRV, the EPA believes
 that these cases would  also be situations where it was
 infeasible to  repair without a process unit shutdown.  Thus,
 the standard already provides the exemption that the
 commenters requested.   The standard does not require routine
 monitoring of  PRV's or  annual compliance demonstrations so the
 EPA does not believe that the provisions pose a major safety
 hazard to monitoring personnel.
      Comment;   Two commenters (A-90-19:  IV-D-77; A-90-20:
 IV-D-19)  recommended that the rule allow installation of a
 rupture  disk under the  pressure relief valve in lieu of
                              5-13

-------
monitoring after a pressure release or monitoring of the
closed vent system.  Several other commenters (A-90-19:
IV-D-69 and IV-D-89; A-90-20:  IV-D-9) suggested that rupture
disks and sensor systems be exempt from the provisions of this
section because these systems detect emissions before they can
be released to the atmosphere.
     Response;   The rule has been clarified to explicitly
exempt PRV's equipped with rupture disks from the follow-up
monitoring requirements of §63.165(b)(2).   For these systems,
the standard requires that a new rupture disk be installed
upstream of the PRV no later than 5 days after the pressure
release, unless the process unit must be shutdown in order to
install the replacement rupture disk.
     Comment;  One commenter  (A-90-19:  IV-D-77) requested
clarification of the requirements for situations in which
relief valves are located on a closed vent system for
protection of the vent system equipment.  The commenter
(A-90-19:  IV-D-77) proposed that instead of being subject to
monitoring, these systems could have a pressure indicator,
located between the rupture disk and the PRV, to indicate the
need for replacement.  The commenter (A-90-19:  IV-D-77) noted
that these valves are not equipped with an isolation valve,
which means the soonest these valves can be repaired would be
at the next process unit shutdown.
     Response;  Rupture disks installed upstream of a PRV are
one of two primary alternatives for control of emissions from
PRV's.  As noted in response to the preceding comment, §63.165
has been revised to clarify that rupture disks are one means
of compliance with the standard.  Thus, the approach proposed
by the commenter is acceptable.   Although it is not clear
from the comment letter whether there is confusion regarding
the applicability of the requirements for PRV's, the EPA would
like to clarify that the provisions in §63.165 only apply to
PRV's that are in organic HAP service (at least 5 weight
percent organic HAP).  Thus, these provisions primarily affect
PRV's on process equipment.
                             5-14

-------
     Comment;  One conunenter  (A-90-19:  IV-D-34) requested
that an exemption be provided for PRVs connected to a common
vent header which discharges to the atmosphere.
     Response;  Section 63.165 (c) exempts PRVs connected to
closed vent systems with control devices from the monitoring
requirement because if the PRV does not reseat properly the
leakage will be controlled.  If the commenter's facility
connects the PRVs to a control device before discharge to the
atmosphere there is no need to monitor the PRV after a
discharge.  However, if the commenter's facility has PRVs
connected to a common header and there is no control before
discharge to the atmosphere, the commenter will either have to
install rupture disks, a control device, or take another
approach toward determining if the PRV has reseated properly.
The EPA does not believe that it would be appropriate to
exempt PRVs from the requirement merely because it would be
inconvenient to comply.
5.1.5  §63.166;  Sampling Connection Systems
     Comment;  Two commenters (A-90-20:  IV-D-19; IV-D-27)
recommended that the definition of closed-loop system be
modified to clarify the intent of the provisions.  The
commenters1(A-90-20:  IV-D-19; IV-D-27)  understanding is that
the intent is to insure that air emissions from sampling are
minimized and the purged material is returned to the process.
One commenter  (A-90-20:  IV-D-27) suggested that the proposed
definition of closed-loop system implies there are no air
emissions and thus there is no difference between closed-loop
and in-situ sampling systems.  The other commenter  (A-90-20:
IV-D-19) thought that §63.166(a) could be interpreted to mean
both the purged material and the sample need to be collected
in a closed-purge, closed-loop, or closed vent system.  The
commenter (A-90-20:  IV-D-19) requested that this section be
clarified.  Another commenter (A-90-19:  IV-D-34) also
requested confirmation that rinsing of sample bottles prior to
sample collection is acceptable as long as the rinsate is
collected and the container is not left open to the
atmosphere.
                              5-15

-------
     Response:  The EPA agrees with the commenters that the
intent of the sampling connection system provisions is to
ensure that purged material is captured and returned to the
process or destroyed, and does not apply to the sample.
Section 63.166 has been clarified regarding the applicability
of the requirements to the sample material.
     The EPA also agrees that it is acceptable to rinse sample
bottles provided the rinsate is collected and properly
recycled or destroyed.  This approach is fully consistent with
the intent of the provisions which is to prevent purging of
process fluids to the ground, sewer drain, or atmosphere.  The
potential for a small amount of emissions during the sampling
procedures is recognized and a zero emissions standard is not
intended.
     Comment:  One commenter (A-90-19:  IV-D-34) requested
that the EPA clarify whether analyzer vents are considered to
be part of the sampling system and subject to controls under
§63.166.  The commenter (A-90-19:  IV-D-34) referenced several
EPA documents and rules to support his view that sampling
connection systems apply at the point the sample is removed
from the process.
     Response:  The EPA agrees that gas streams exiting an
analyzer are not considered to be subject to the provisions of
§63.166.  The commenter is correct in noting that the focus of
this provision is at the point where samples are removed from
the process.  The EPA believes that the question has arisen
due to the lack of clarity regarding the meaning of the term
analyzer vent.  In particular, the term "analyzer vent" has on
occasion been used to refer to the gases purged through a
sample manifold system.  In these systems, the analyzers
remove a sample of the gas from the manifold.  The provisions
in §63.166 would apply to the gas flow through the manifold,
but would not apply to the gases exiting the analyzer.
     Comment:  Two commenters (A-90-19:  IV-D-34; IV-D-54)
suggested that the EPA provide examples, in the form of
diagrams or drawings, showing acceptable sampling systems,
including commercially available systems.  The commenters
                             5-16

-------
 (A-90-19:  IV-D-34; IV-D-54) indicated that illustrations are
 necessary because the definitions in the proposed rule are
 confusing.
     Response;  The EPA agrees that drawings could be helpful
 to supplement the definitions given in the regulation.
 Although the commenters did not clearly indicate whether they
 wanted the illustrations to appear in the regulation or the
 BID, the EPA feels that the BID would be the more appropriate
 place.  An illustration has been provided in the appendix of
 this document.
     Comment;  One commenter (A-90-20:  IV-D-19) recommended
 that §63.166(c) be expanded to exempt sampling systems without
 purges from the equipment standard requirements in paragraphs
 (a) and (b).  The commenter (A-90-20:  IV-D-19) also
 recommended that non-routine grab samples taken during process
 upset conditions be exempted.
     Response;  Section 63.166(c) has been revised to also
 exempt sampling systems without purges from the requirement to
 use closed-loop, closed-purge, or a closed vent system.  The
 change was made because it is possible in some cases to design
 sampling systems to collect samples without purging the sample
 line.  It is expected that the owner or operator of the source
 will be able to show that the system is operated without
 purges or why it is infeasible to purge materials through this
 sampling system.
     The suggestion that non-routine grab samples also be
 excluded from the provisions in §63.166 was not adopted.  The
 EPA believes that the owner or operator should include
 activities such as this in the startup, shutdown, and
 malfunction plan.  The provisions in subpart H are intended to
 apply during periods of normal operation and not during
 malfunctions and process upsets, which should be addressed in
 the source's startup, shutdown, and malfunction plan contained
 in §63.6(e)(3)  of the General Provisions.
     Comment;   One commenter (A-90-19:  IV-D-60) requested
that equipment in heavy liquid service, that has a
concentration of less than 500 ppm in the line, be exempted
                             5-17

-------
from the sampling connection system requirement.   The
commenter (A-90-19:  IV-D-60) suggested that this exemption
would be appropriate because some materials have extremely low
vapor pressures and, therefore, essentially no emission
potential.
     Response;  If the commenter is referring to a stream
composition less than 500 ppm of HAP, the provisions would not
apply.  In order for the provisions to apply, the equipment
must be in organic HAP service, which is defined as:
     In organic HAP service means a piece of equipment
     either contains or contacts a fluid (liquid or gas)
     that is at least 5 percent by weight of total
     organic HAP's. . .
On the other hand, if the commenter is referring to the
concentration of the vapor above the liquid, the EPA does not
agree that it would be appropriate to exempt equipment in
heavy liquid service from this requirement.  Heavy liquid
streams have the potential to emit VOC's and organic HAP's to
the atmosphere, particularly from purged materials that are at
elevated temperatures or materials purged to sewer drains.
Since the requirements for sampling connection systems allows
the use of closed-purge systems as well as closed-loop
sampling, the EPA believes the standard is achievable for
equipment in heavy-liquid service.
5.1.6  §63.167;  Open-ended Valves or Lines
     Comment;  One commenter (A-90-19:  IV-D-34) requested
that an exemption be provided for emergency shutdown systems,
which are designed to open automatically during process
upsets.  These automatically opening vent lines must never be
closed even with a second valve.
     Response:  The EPA agrees that automatically opening vent
lines which are part of an emergency shutdown system should
not be required to add a second valve or cap.  It was also
determined that the requirements for block and bleed systems
were not appropriate.  Section 63.167(d) was, therefore, added
to the final rule to address a potential safety hazard.
     Comment:  One commenter (A-90-19:  IV-D-60) requested
that equipment in heavy liquid service be exempt from this
                             5-18

-------
 requirement due to the very low potential for emissions.  The
 commenter  (A-90-19:  IV-D-60) reported that a Method 21 survey
 of their caprolactam plant showed the majority of 5,000 open
 ended  lines had no detectable emissions readings and the
 single highest concentration recorded was 21 ppm.
     Response;  According to the analysis that accompanied a
 previous equipment leak standard (40 CFR part 60, subpart W,
 [48 FR 48328]), these controls are cost-effective and it is
 common practice in the industry to cap lines.
 5.1.7  363.168;  Valves in Gas/Vapor Service and in Light
 Liquid Service
     Comment;  One commenter (A-90-19:  IV-D-92) suggested
 that the subpart should include a random 200-valve test as an
 alternative standard for valves.  The commenter  (A-90-19:
 IV-D-92) recommended that such an option could allow units
 that stayed below 2 percent leaking valves  (where a leak is
 defined as 2000 ppmv) for 2 years to randomly test 200 valves
 or 10  percent of the valves annually.  Whenever the unit
 exceeded 2 percent leaking valves, the owner would be required
 to resume quarterly monitoring.  The commenter  (A-90-19:
 IV-D-92) stated that this option would provide an incentive
 for compliance.
     Response:  Although certain aspects of the commenter's
 suggestion have merit, it must be noted that the numbers
 mentioned are significantly less stringent than the levels
 agreed to by the negotiating committee.  For example, after
 the first year, the definition of leak becomes 500 ppmv, not
 2,000 ppmv as suggested by the commenter.  Additionally, the
 commenter indicated that quarterly monitoring would be
 appropriate for units exceeding the 2 percent leakage rate,
 yet the original agreement was for QIP or monthly monitoring
 for this situation.
     The EPA also notes that the committee did consider
 several options for random-sampling alternatives, all of which
were rejected due to difficulty in determining whether samples
were truly random.
                              5-19

-------
     Comment;  Two conunenters (A-90-19: IV-D-32; A-90-20:
IV-D-3) requested that the EPA clarify §63.168(e)(l) by
specifying how leaks that recur within 90 days of repair
should be considered in the calculation of percent leaking
valves.  The conunenters (A-90-19: IV-D-32; A-90-20:  IV-D-3)
questioned whether the leak should be  (1) treated as a new
leak;  (2) treated as a leak for which the initial attempt at
repair failed; or (3) put on the list for repair during the
next process unit shutdown.  The commenter (A-90-19: IV-D-32;
A-90-20:  IV-D-3) suggested that either  (1) the recurrence be
treated as a new leak from a repair standpoint, but not be
counted in the percentage of leaks; or (2) the equipment be
put on the nonrepairable list and repaired at the next
shutdown.
     Response;  Section 63.168(e)(l) was clarified to specify
that the calculation of percent leaking valves is based on the
number of valves determined to be monitored during the
periodic monitoring.  Thus, if the process unit is on a
quarterly monitoring schedule the follow-up monitoring at
90 days would be conducted as part of the routine quarterly
screening and if the valve is found to be leaking again would
be counted as a leak.  If the process unit is on a semiannual
or annual monitoring schedule, the valve would have to be
repaired but not be counted in the calculation of percent
leaking valves.  If the owner or operator determines that the
valve must be removed in order to repair it in the shop, then
the valve would be put on the nonrepairables list.
     Comment;  One commenter  (A-90-20:   IV-D-19) requested
that the EPA clarify the monitoring schedule if a new valve  is
installed.  This commenter (A-90-20:   IV-D-19) noted that
§63.168(f) specifies procedures for repairing leaking valves,
but does not provide for repairs completed by replacing the
valve.  The commenter (A-90-20:  IV-D-19) noted that it would
be unmanageable if the valve must be kept on a separate
monitoring schedule from the other valves in the process unit.
     Response;  As discussed earlier in  response to a similar
comment on the pump standard, subpart  H  specifies the
                              5-20

-------
monitoring frequency and the phases of the standard on a
process unit basis.  The standard does not establish
monitoring frequency for individual items of equipment.  The
EPA agrees with the commenter that it would be unmanageable to
have individual valves on different monitoring schedules.  To
minimize possible implementation problems, the EPA will
explain in enabling materials, and inspection manuals that the
phases of the pump and valve standards are determined on a
process unit basis and addition of a new valve or pump does
not alter the phase the equipment is considered to be in or
the monitoring frequency of the equipment.
     The EPA would also like to clarify that §63.168(f)
provides a partial list of actions that can be taken as a
first attempt at repair.  These examples were included at the
request of some Committee members because of their experiences
in implementing the existing equipment leak standards in
40 CFR parts 60 and 61.  The intent of including §63.168(f)
was to illustrate the extent of actions necessary to comply
with the first attempt at repair.  Valve replacement was not
included because the Committee did not envision this action as
being a first attempt at repair measure.  In many cases, valve
replacement would require a process unit shutdown or bypassing
of equipment and draining of process fluids from the lines in
the affected area.
     Comment;  One commenter (A-90-20:  IV-D-19) argued that
the standard should not require post-repair monitoring of
unsafe-to-monitor valves and connectors.  The commenter
(A-90-20:  IV-D-19) noted that it is not feasible to remonitor
the equipment within 3 months as required by §63.168(f)(3).
     Response;  To ensure that there is no possibility of
misunderstanding the requirements for unsafe-to-monitor
valves, the provisions in §63.168(h) have been edited to
exempt these valves from the requirements in paragraphs  (b)
through (f)  of §63.168.
     Comment;  A number of commenters (A-90-19:  IV-D-33;
IV-D-60; IV-D-73; IV-D-86) (A-90-20:  IV-D-20) recommended
that the proposed rule be modified to allow facilities to
                             5-21

-------
begin Phase III on the applicability date of the rule.  The
commenters (A-90-19:  IV-D-60; IV-D-73; IV-D-86) (A-90-20:
IV-D-20) submitted that disallowing this would penalize
facilities that have established low leak rates or that have
implemented the proposed rule before the required
applicability date.
     The commenters (A-90-19:  IV-D-60; IV-D-73; IV-D-77;
IV-D-86) (A-90-20:  IV-D-20) suggested a variety of criteria
for entering Phase III, including allowing the source to
decide when it is appropriate and demonstrating that the
required percent leaking criteria was achieved during the two
most recent monitoring periods.  Specifically, one commenter
(A-90-19:  IV-D-73) suggested that if an owner or operator can
demonstrate at any time that the source qualifies for reduced
monitoring frequency,  the source should be allowed to adopt
the reduced monitoring frequency.  Another commenter  (A-90-19:
IV-D-14) recommended that the source be required to
demonstrate that the criteria were achieved during the two
most recent monitoring periods.  Another commenter (A-90-19:
IV-D-77) thought that the rule should allow the owner or
operator to elect the monitoring frequency most appropriate to
the source's current status because the'necessary records may
not have been retained.
     Response;  The EPA agrees that the final rule should
allow owners or operators the flexibility to initiate
Phase III at anytime,  and it was intended that this option
would be available.  Subpart H has been revised to clarify
this point.   This clarification does not, however, allow an
owner or operator to elect to use reduced monitoring
frequencies without Method 21 data to document achievement of
lower leak rates for the required periods.
     Comment;  One commenter  (A-90-19:  IV-D-60) requested
that the rule allow the owner or operator the option  of
calculating percent of leaking valves on a plant-wide basis or
a process-unit basis.   The commenter (A-90-19:  IV-D-60) also
requested that the rule allow the owner or operator the option
of grouping units that are in similar service.  The commenter
                              5-22

-------
 (A-90-19:  IV-D-60) maintained that this would assist in
providing meaningful leak rate data and would still meet the
intent of the rule.
     Response; The EPA is uncertain just what the commenter
meant by meaningful leak rate data.  In Phase III of the valve
standard, the monitoring frequency is determined by the
percent leaking valves.  Since the number of valves in a
typical SOCMI process unit is quite large (i.e., several
hundred to thousands), the variability of the estimate should
be small and it should not be necessary to combine data from
several units to obtain a reliable estimate of performance.
It should also be noted that the provisions in the valve QIP
 [§63.I75(e)(2)] allow pooling of performance data for purposes
of identifying measures to improve performance.  Therefore, if
the commenter's concern was with obtaining better data on
performance for certain operating conditions, the standard
already allows this.
     Comment;  One commenter (A-90-20:  IV-D-19) recommended
that the EPA reconcile two conflicting methods for determining
whether excessive leaks trigger additional requirements for
process units following annual monitoring.  The commenter
 (A-90-20:  IV-D-19) explained that §63.168(d)(1} specifies
that "process units with 2 percent or greater leaking valves,
calculated as a rolling average of 2 consecutive periods"
shall either go to a QIP or implement monthly monitoring.  For
a process unit in annual monitoring, this conflicts directly
with §63.168(e)(2), which states that "the percent leaking
valves shall be calculated as...an average of any three out of
four consecutive monitoring periods for annual monitoring
programs".   Even though a process unit has reached an annual
monitoring program, it may still obtain two consecutive
periods of greater than 2 percent leaking valves.  For
example, a process unit with an annual monitoring schedule
could monitor for four quarters and have the following percent
leaking valve numbers: 1 percent, 1 percent, 3 percent and
5 percent.   According to paragraph (e)(2), any three out of
four of these periods could be used, so the unit would have an
                             5-23

-------
average of 1.67 percent leaking valves [(1% + 1% + 3%)/3 =
1.67%], and would need to drop to quarterly monitoring, as per
paragraph (d)(2).  But paragraph (d)(1)  says that the 2
consecutive monitoring periods greater than 2 percent put the
process unit into either monthly monitoring or a QIP.  Which
is correct?  Since paragraph (e)(2) matches the method given
in the preamble for determining monitoring frequency, it is
assumed that paragraph (d)(1) is in error.
     Response;  In drafting the provisions of §63.168(d), it
was assumed that a source with an annual monitoring program
would not have quarterly monitoring data and, thus, the
scenario presented by the commenter could not arise.  Since
this comment suggests that it is possible for the provisions
in §63.168(d) to be interpreted in a manner that appears to
conflict with the provisions in §63.168(e), the provisions in
§63.168(d) have been edited to remove this possibility.
Section 63.168(d)(l) now refers to the percent leaking
calculated according to §63.168(e).
     Comment;  One commenter (A-90-19:  IV-D-73) recommended
that the difficult-to-monitor criteria be revised to not
require elevation of monitoring personnel above support
surfaces that are accessible only by fixed ladder.  The
commenter (A-90-19:  IV-D-73) noted that it is unsafe to carry
stepladders up fixed ladders.  Another commenter  (A-90-19:
IV-D-69)  recommended that the criteria for both "difficult-to-
monitor valves" and "inaccessible connectors" should be the
same - i.e., equipment is no more than 2 meters above a
support surface.  The commenter (A-90-19:  IV-D-69) noted that
a valve and a connector next to each other would be handled
differently under the proposed provisions.
     Response:
     A clarification has been added to the equipment leak
valve provisions to specify that valves are considered
"difficult-to-monitor" if they are more than 2 meters above a
support surface or the elevation of personnel on  support
surfaces can not be conducted safely at anytime.
                              5-24

-------
     Comment;  One commenter  (A-90-19:  IV-D-73) recommended
that §63.168(1)(2) be deleted.  The commenter  (A-90-19:
IV-D-73) stated that even in new facilities an owner or
operator cannot ensure that all valves will be readily
accessible because placement is dictated by process
requirements.
     Response:  The EPA contacted the commenter to determine
the reasons this requirement had not been an issue under the
benzene equipment leaks standard in 40 CFR 61, subparts J and
V.  The commenter noted that there are major differences
between the number of units and the amount of equipment
affected by the HON and by the benzene equipment leak NESHAP.
Due to the greater magnitude of the HON, it is just not
possible for existing sources to reconfigure process equipment
to accommodate this standard.  This commenter also pointed out
that the SOCMI equipment leaks NSPS (40 CFR 60, subpart W)
placed no limit on the number of difficult-to-monitor valves
in existing units affected due to modification or
reconstruction and allowed new units to have up to 3 percent
difficult-to-monitor valves.  Thus, §63.168(i)(2) was revised
to be consistent with the provisions in the NSPS.
     Comment:  One commenter  (A-90-19:  IV-D-89) questioned
the benefit of monitoring "leakless" valves with the same
frequency as all other valves.  The commenter  (A-90-19:
IV-D-89) suggested that no more than quarterly monitoring
should be required for these valves.
     Response;  During negotiations, the committee discussed
providing special provisions for different types of valves and
concluded that this additional complexity was not useful.
Specifically, if a process unit had a large number of leakless
valves, it probably would not have 2 percent or greater
leaking valves and thus not be subject to monthly monitoring
or QIP requirement.  It is very likely the process unit could
qualify for semiannual or annual monitoring frequency.
Conversely, if there were only a few leakless valves, the
burden of identifying them and treating them differently would
undoubtedly exceed the monitoring burden.
                              5-25

-------
     Comment;  One cor  nter (A-90-20:  IV-D-19) suggested
that un:  fe-to-monitcr and difficult-to-monitor valves should
be excluded from the percent leaking valves calculation.   The
coinmenter (A-90-20:  IV-D-19) noted that because of the
infrequent monitoring performed on these valves, unsafe-to-
monitor and difficult-to-monitor valves should not be counted
in the percent leaking valves calculation given in
§63.168(e)(1).  If all valves in this group are monitored in
one period,  the results could be very skewed in one direction
or another,  giving an unrealistic picture of the remaining
valves in the process unit.  These valves should be excluded
from both the number of leaking valves (VL) and total valves
(Vij)  terms.    The commenter  (A-90-20:  IV-D-19) also thought
that the verbiage for how to calculate VL in paragraph
(e)(3)(i) is hard to follow; an equation would be more
helpful,  along with some rewording of paragraphs (e)(3)(i) and
(e)(3)(ii),  which address nonrepairable valves.
     Response:  The equation as drafted in the proposed rule
calculated the percent leaking in the population of valves
monitored during that particular monitoring cycle.  The EPA
doubts that the commenter's concern is likely to arise in
practice since the commenter can schedule monitoring of valves
designated as "difficult-to-monitor" or "unsafe-to-monitor" to
avoid this problem.  There is no requirement to do this
monitoring during a periodic monitoring cy^le and in fact it
is assumed that this monitoring would not be conducted during
routine operations.  Thus, it seems highly unlikely for a bias
to be introduced.  Moreover, since these valves are, by
definition,  unlikely to be moved frequently there is no reason
to believe that these would have higher leak frequencies than
the accessible valves.  The EPA also suspects thar there may
have been some confusion on the part of the commenter between
the terms "nonrepairable" and "difficult-to-monitor" and
"unsafe-to-monitor".   Thus, the commenter's suggestion to
edit §63.168(e)(3) was not adopted; however, guidance material
on the standard will include an equation for §63.168(e)(3) to
assist with implementation.
                             5-26

-------
 5.1.8   §63.169:  Pumps. Valves. Connectors, and Agitators in
 Heavy Liquid Service; Instrumentation Systems; and Pressure
 Relief  Valves in Liquid Service
     Comment;  One commenter  (A-90-19:  IV-D-60) requested
 that the rule be modified to  address situations where a
 potential leak is observed but subsequent monitoring shows
 repair  is not required.  The  commenter  (A-90-19:  IV-D-60)
 expressed concern that the proposed rule did not indicate that
 remonitoring is not required.  The commenter (A-90-19:
 IV-D-60) suggested that in such cases the equipment should be
 exempt  from remonitoring for  90 days.
     Response; The EPA assumes that the remonitoring referred
 to by the commenter would be  the result of the reappearance of
 drips or other visible signs  of seal leakage.  Visible leakage
 from pump seals is generally  indicative of seal wear and to
 prevent major seal failures,  the seals should be repaired soon
 after the leakage is initially detected.  If the situation
 described by the commenter is such that the. pump is not
 repairable, the owner or operator can put the pump on the
 nonrepairables list and repair it at the next process unit
 shutdown.  In such cases, the pumps would be exempt from
 remonitoring.
     Comment;  One commenter  (A-90-19:  IV-D-69) thought that
 for all components the rule should provide that if a potential
 leak is discovered the owner  or operator has the option to
 assume that it is a leak and  repair it and not have to monitor
 to confirm the leak.  The commenter thought that it was
 illogical to provide this provision only for instrumentation
 systems.
     Response;  Special provisions were developed for
 instrumentation systems because of the physical difficulty of
monitoring individual components in these systems and the
nature of these systems would allow confirmation of successful
repair as well as indication of the presence of a leak (such
as a change in pressure or flow rate).  Other components
subject to this rule do not share these characteristics, and
there would be no way to confirm that the leak was repaired.
                              5-27

-------
     Comment;  One commenter (A-90-19:  IV-D-60)  requested
that these provisions not apply to equipment where it can be
demonstrated that the equipment would never be considered to
be leaking when monitored by Method 21.
     Response;  The EPA believes that it is appropriate to
retain the requirements to repair equipment with indications
of leakage and the requirement is not burdensome.  In many
cases, these indications of potential leaks are indicative of
pending major seal failure.  Although equipment in heavy
liquid service has much lower emission rates than equipment in
light liquid or gas service, losses of process fluids to the
environment should be minimized because it will ultimately be
lost to the atmosphere or could contribute to groundwater
contamination.
5.1.9  §63.170;  Product Accumulator Vessels
     Comment:  Several commenters (A-90-19:  IV-D-50; IV-D-74;
IV-D-77)  favored regulating product accumulator vessels under
the provisions for process vents in subpart G and indicated
that §63.170 should be eliminated.
     Response:  As discussed extensively in volume 2D of the
BID, some of the equipment previously covered by this term is
considered to be a process vent.  The final standard has
eliminated this overlap and the provisions in §63.170 now
apply only to surge control vessels and bottoms receivers.
These vessels do not meet the definition of a process vent (or
a storage vessel) and have intermittent releases only.  The
EPA, therefore, concluded that retaining this equipment in
subpart H would be consistent with the negotiated agreement.
     cqmjnent;  one commenter (A-90-19:  IV-D-108) argued that
by incorporating the negotiated equipment leak rule into the
HON, pharmaceutical manufacturers have become subject to a new
and different standard for the vent emissions from product
accumulators.  The commenter (A-90-19:  IV-D-108) urged the
Agency to regulate these product accumulator vent emissions
under the vent standards for the pharmaceutical source
category and not subject a source to overlapping or
contradictory standards.
                             5-28

-------
     Other commenters  (A-90-19:  IV-D-7; IV-D-39) presented
similar arguments that the negotiation  did not include point
sources such as PAV's and that the negotiations did not
address what performance would be achievable for
pharmaceutical processes.  Another commenter (A-90-19:
IV-D-27) thought that it was unnecessary and undesirable to
regulate PAV's in the equipment leak rule since these are more
appropriately addressed under process vents.  This commenter
(A-90-19:  IV-D-27) also argued that removing PAV's from
subpart H would address the inequity which exists for batch
processes in the proposed rule.
     Response:  The EPA believes that several clarifications
to the final rule have addressed the commenters' concerns.
First, as noted above, the overlap between process vents and
equipment included in the proposed definition for PAV's has
been eliminated.  Of the original items included in the
definition of PAV's, subpart H now only establishes
requirements for surge control vessels  and bottoms receivers.
Second, the applicability for the non-SOCMI processes has been
separated from that for the SOCMI processes.  Subpart I now
has the applicability for the non-SOCMI processes and
subpart F has the applicability for SOCMI processes.  Thus,
for sources subject to subparts H and I, there are no control
requirements for process vents.
5.1.10  S63.171;  Delay of Repair
     Comment;  One commenter (A-90-19:  IV-D-87) argued that
the option of using delay of repair should include the
examples cited in the proposed standard as well as an
evaluation of the potential to cause any adverse effects to
human health and the environment.
     Response;  Provisions for delay of repair have been a
feature of the equipment leak standards in 40 CFR parts 60 and
61 since the beginning of the EPA's program.  The EPA has
provided this extension because it would be counterproductive
to establish a requirement that would result in release of
more emissions to repair the leaking component than would
occur if the component was left unrepaired.  The commenter's
                             5-29

-------
(A-90-19:  IV-D-87) suggestion that the potential for adverse
health and environmental impacts also be a criterion for delay
of repair is not appropriate for this rule.  The residual risk
standards to be established under section 112(f) would be more
appropriate than this standard.  It should also be noted that
units that handle the acutely toxic HAP's, such as phosgene,
are designed to permit rapid shutdown of the equipment on any
indication of a leak and immediate repair.
     The EPA believes that the provisions in §63.171 do
provide delay of repair for any type of equipment.  So, the
EPA is not certain what the nature of the commenter's
(A-90-19:  IV-D-69) concern is.
     Comment:  One commenter (A-90-20:  IV-D-19) recommended
expanding the delay of repair provisions to include replacing
any seal system with one that is expected to provide better
performance.  The commenter (A-90-20:  IV-D-19) observed that
the plant may wish to replace DMS systems or sealless pumps
with more efficient systems, and this replacement should also
be allowed.  The commenter  (A-90-20:  IV-D-19)  also noted that
the replacement of single seal systems is generally not
required so much as desired.  The commenter (A-90-20:
IV-D-19) thought that documentation of procedures describing
how a replacement was determined to be actually required would
be lengthy and burdensome.
     Response:  The provisions in §63.171(d) were revised to
allow delay of repair for systems expected to achieve better
performance.  This suggestion was considered appropriate and
provided delay of repair conditions equivalent to those
provided for sources subject to the provisions of §63.176.
The EPA would also like to clarify that the necessary
documentation for use of this provision is merely that the
pump cannot be repaired by normal procedures.  While the EPA
understands that there may be a number of options available
and the owner or operator may elect seal replacement,  it would
be inappropriate to edit the language as suggested since that
could result in feasible repairs not being done.
                             5-30

-------
     Comment;  One commenter  (A-90-19:  IV-D-33) recommended
 revising  §63.171(e) to apply  to all equipment, not just
 valves, to recognize that stocks of some of the specialized
 equipment components may be depleted and not be available on
 short notice.
     Response;   The equipment leak provisions have been
 revised to apply this allowance to connectors.  This allowance
 was not extended to other equipment because the quantity of
 other types of equipment (i.e., pumps, agitators or
 compressors)  used at a facility is much smaller than the
 quantity  of valves and connectors.  Therefore, the possibility
 of the quantity needed to be  stocked being incorrect is
 remote.
     Comment;  One commenter  (A-90-19:  IV-D-33) recommended
 that delay of repair provisions for pumps [§63.17l(d)(2)] not
 require the repair to be completed within 6 months.  The
 commenter (A-90-19:  IV-D-33) argued that the time restriction
 could cause problems at plants that have infrequent
 maintenance shutdowns.
     Response;  The provisions of §63.171 allow delay of
 repair for three situations:  (1) where a process unit shutdown
 is required;  (2) where the equipment is isolated from HAP
 service;  and  (3) where a better performing seal system is
 going to  be installed.  The 6-month time limitation only
 applies in the last case.  If repair is technically infeasible
 without a process unit shutdown, the delay is until the next
 process unit shutdown, which could be a delay of more than 6
 months or less than 6 months depending on operations of the
 unit.  Obviously, if the pump is removed from organic HAP
 service the delay can be as long as the owner or operator
 wishes to keep the pump out of organic HAP service,  since
 there are a number of options available in addition to the
 delay provided by §63.171(d), the EPA does not believe that
 the time restriction should be removed.  If the time limit on
 this delay were removed, it is conceivable that someone could
use this provision to avoid repairing a leaking pump
altogether.
                              5-31

-------
5.1.11  §63.172:  Closed-vent Systems and Control Devices
     Comment:  Several conunenters (A-90-19:  IV-D-33; IV-D-56;
IV-D-73; IV-D-77) (A-90-20:  IV-D-19) advised that valves and
connectors in closed vent systems should be subject to the
same standards as regular valves and connectors, which allow
delay of repair until the next process unit shutdown if such
repairs cannot be made without a shutdown.  One commenter
(A-90-19:  IV-D-73)  declared that the delay of repair
provisions in §63.120(f)(2) should be incorporated into
§63.172.  One commenter (A-90-19:  IV-D-77;) also suggested
that provisions should be added to allow less frequent
monitoring in systems with low percentages of leaking
components.
     One commenter (A-90-19:  IV-D-73) requested that the EPA
conduct a comprehensive review of subparts G and H for
requirements that could apply to the same equipment if they
were part of a common control system.  The commenter (A-90-19:
IV-D-73) noted that the requirements should be consistent, and
contradictions should be eliminated.  The commenter (A-90-19:
IV-D-73) recommended that the exclusions in subpart G for
bleeds, drains, pressure vacuum vents, etc. be incorporated
into subpart H.  The commenter (A-90-19:  IV-D-*73) also
requested that §63.160(d)(3) include an override of subpart G
if delay of repair provisions are applied to closed vent
systems.  A similar comment was made by another commenter
(A-90-19:  IV-D-97)  who recommended that all the closed vent
system provisions be consolidated in subpart H.  The commenter
(A-90-19:  IV-D-97)  noted that this change would reduce any
confusion over the requirements and would reduce the
recordkeeping and reporting costs.
     Response:  The EPA agrees with the commenters1
suggestions that it would be appropriate to have a consistent
set of provisions for closed vent systems in the rule.  A
uniform set of provisions for closed vent systems will benefit
both State and Federal enforcement programs and industry by
both reducing review time and complexity of record systems.
Because subpart G also included requirements for inspections
                             5-32

-------
 of  equipment other than closed vent systems, the closed vent
 system provisions in subpart G were not consolidated  into
 subpart H.  The  final rule now has the same requirements for
 closed vent systems in subparts G and H.  The EPA believes
 that this approach provides the consistency requested by the
 commenters.
     The EPA also reevaluated the provisions requiring annual
 Method 21 monitoring of closed vent systems.  Closed  vent
 systems in chemical plants and refineries are constructed of
 piping and connections and are operated at low pressures or
 under vacuum.  An assessment of recent data and experience
 from implementation of existing standards under 40 CFR part 60
 and part 61 showed that only rarely are leaking connectors and
 other equipment  identified through the annual Method  21
 inspections of closed-vent systems.  As discussed in  the
 preamble to the  proposed rule (57 FR 62666 and 57 FR  62676),
 connectors have  very low leak frequencies and once leak tight
 they remain leak tight.  Consequently, the final rule only
 requires an initial Method 21 demonstration that all
 connections and  other equipment in closed vent systems are
 operated with instrument readings less than 500 ppm and annual
 inspections for  indications of leaks (visual, olfactory, or
 audible).   The  EPA believes that this requirement along with
 the requirement  for flow indicators or car seals on by-pass
 lines that could divert emissions from the control device to
 the atmosphere will ensure emissions are controlled as
 required, while  also minimizing unproductive effort.
     rodent;  one commenter (A-90-19:  IV-D-75) indicated
 that the requirements for controlling certain equipment with a
 closed vent system and a control device results in the
 equipment being  subject to the process vent provisions in
 subpart G.  The  commenter (A-90-19:  IV-D-75) recommended that
 the final rule should provide the owner or operator the option
 of complying with either the requirements in subpart  G or
 those in subpart H.
     Response;   The EPA suspects that there may be a
misunderstanding of the meaning of the term "process  vent".
                             5-33

-------
A process vent means a gas stream that is continuously
discharged during the operation of the unit from an air
oxidation reactor, other reactor, or distillation unit within
a SOCMI chemical manufacturing process unit.  Process vents
include vents from distillate receivers and product
separators.  Process vents include gas streams .that are
discharged directly to the atmosphere and gas streams
discharged to the atmosphere after diversion through a product
recovery device.  Thus, it is not possible for equipment
subject to subpart H to be considered subject to the
provisions in subpart G merely because it has been connected
to a closed vent system.  For something to be subject to the
process vent provisions it must meet the definition of a
process vent.
     Comment;  One commenter (A-90-19:  IV-D-33) argued that
the control device performance should not be evaluated on the
basis of all organics routed to it, just HAP control
performance.
     Response;  The provisions in §63.172 were edited to allow
the owner or operator to demonstrate the performance based on
either organic HAP's or VOC .
5.1.12  §63.173;  Agitators in Gas/Vapor Service and in Light
Licruid Service
     Comment;  One commenter (A-90-19:  IV-D-56) alleged that
the proposed rule expanded the list of equipment subject to
the rule by adding agitators and instrumentation systems.  The
commenter (A-90-19:  IV-D-56) opposed this extension because
it goes beyond the negotiated agreement without providing
reasonable justification.
     Response;  The Committee developed the provisions for
instrumentation systems and agitators.  The preamble to the
proposed rule describes the factors considered by the
Committee and the reasons for including this equipment in the
standard (see December 31, 1992, Federal Register [57 FR
62080]).  The commenter (A-90-19:  IV-D-56) may have
misinterpreted discussions comparing the negotiated rule to
existing equipment leak standards in 40 CFR parts 60 and 61.
                             5-34

-------
Agitators are not subject to the provisions in the earlier
standards, and the Committee elected to add this equipment to
the scope of the negotiated rule.
     Comment;  Several commenters  (A-90-19:  IV-D-73; A-90-20:
IV-D-9; IV-D-12)) requested that agitators equipped with
double seals be exempt from monitoring, as are pumps.  Another
commenter (A-90-20:  IV-D-19) suggested that exemptions from
routine monitoring be provided for better agitator designs and
provisions for unsafe to monitor equipment be added.
     Response:  The final standard for agitators includes
provisions for agitators equipped with dual seals and for
agitators equipped with a closed vent system.  Agitators with
no externally actuated shaft are exempt from the monitoring
requirements of the standard.  Since the commenter  (A-90-20:
IV-D-19) did not provide examples of situations where
monitoring of the agitator would be unsafe, no provisions were
added to exempt these situations.
5.1.13  §63.174;  Connectors in Gas/Vapor Service and in Light
Liquid Service
     Comment;  One commenter (A-90-19:  IV-D-56) requested
that the final rule include provisions allowing a facility to
make connectors leak-proof by welding them and, therefore, to
receive credit in the calculation of percent leaking.
     Response;  The negotiated rule does provide credit for
removing connectors from a process.  These provisions are in
§63.174(i) of subpart H.
     Comment:  One commenter (A-90-20:  IV-D-19) objected to
the proposed requirement in §63.174(b)(4) arguing that the
requirement to monitor welds used to reduce the number of
connectors in a process unit is unnecessary and requires
regulation before applicability.  The commenter (A-90-20:
IV-D-19) noted that industry practice when welding any type of
equipment is to test the weld integrity before placing the
equipment back in service.  The proposed requirement would not
provide for any additional protection to the environment, and
would greatly increase the recordkeeping burden of a process
                              5-35

-------
unit.  In addition, this section requires both testing and
recordkeeping before any valid applicability date.
     Response:  The provisions in proposed §63.174(b)(4) were
intended to apply to the optional credit for removed
connectors, and not to create an additional recordkeeping
burden on sources that did not elect to use the credit.  The
provision allowed credit back to the date of proposal because
at the time of the negotiation some Committee members
advocated providing the maximum opportunity to generate
credits.  As this comment showed it was possible to read the
proposed provisions in §63.174(b)(4) as not being voluntary,
the EPA redrafted this provision and moved it after the
calculation of percent leaking connectors.  It is hoped that
these editorial changes will make the provision clearer.
     Comment:  One commenter (A-90-19:  IV-D-60) requested
that the rule provide the option of calculating percent
leaking on a plant-wide basis or a process unit basis.  The
commenter (A-90-19:  IV-D-60) also requested that the rule
allow the owner or operator to group process units that are in
similar service for calculating percent leaking.  The
commenter (A-90-19:  IV-D-60) thought this would assist in
providing meaningful leak rate data.
     Response;  This comment is addressed under the valve
standard.
     Comment;  One commenter (A-90-19:  IV-D-68) argued that
the proposed LDAR requirement for connectors should be
eliminated from the final rule because the proposed
requirement is above the floor for equipment leaks and will
result in negligible emission reductions.  The commenter
(A-90-19:  IV-D-68) also argued that this program would
require 4,000 to 5,000 man-hours to implement in a process
unit with about 50,000 components.  Therefore, the commenter
(A-90-19: IV-D-68) concluded that this is not a cost-effective
approach to emissions reduction.  For similar reasons, another
commenter (A-90-19:  IV-D-77) recommended that only an annual
inspection for leaks based on visual, auditory, or olfactory
detection should be required for connectors.
                             5-36

-------
     Response;  The EPA does not agree with the commenter's
 (A-90-19:   IV-D-68) view that a LDAR program for connectors is
 inappropriate and  is not a cost-effective means of emissions
 reduction.  The commenter  (A-90-19:  IV-D-68) did not provide
 the basis  for the  emission estimates used in concluding that
 the LDAR program for connectors was not cost-effective.  The
 EPA believes that  it is important to include process equipment
 connectors  in the  LDAR program because emissions from these
 connectors  can be  significant.  The revised SOCMI average
 factors show that  the factor for connectors is one-half to
 one-third of the factors for valves in light liquid and gas
 service.  Because  of the large number of connectors in process
 units, connector emissions could easily exceed emissions  from
 valves and  pumps.  In fact for the number of components
 reported by the commenter  (A-90-19:  IV-D-68), the revised
 SOCMI average factors indicate that connectors contribute
 roughly 55  percent of total emissions and valves contribute 40
 percent.  While the average factors may not be indicative of
 emission rates for the commenter's (A-90-19:  IV-D-68) units,
 they do indicate that on a national basis it is important to
 consider control measures  for connectors.  The EPA considers
 the negotiated rule, as well as the connector LDAR program, to
 be a cost-effective means of reducing emissions from equipment
 leaks.  Since the  standard allows less frequent monitoring for
 better performing  units, the EPA does not believe that the
 provisions  will impose unproductive costs on units that
 perform better than the average units.
     The suggestion that the connector LDAR program be
 replaced with an annual inspection program for indications of
 leaks was rejected for the same reason.  The EPA believes that
 the program can be cost-effective.  Additionally, it would not
 be consistent with the negotiated agreement to remove the
 connector LDAR program without providing an equivalent
 reduction from other items of equipment.  The second commenter
 (A-90-19:   IV-D-77) did not suggest any substitute control
measures or provide reasons for the view that emissions were
trivial.
                             5-37

-------
     Comment;  One commenter (A-90-19:  IV-D-77)  requested
that the final rule allow the owner or operator to adopt a
reduced monitoring frequency if it can be demonstrated that
the source qualifies for this frequency.  Another commenter
(A-90-20:  IV-D-14) suggested that facilities be allowed to
implement the less frequent monitoring provisions of
§63.174(b)  provided that the percent leaking connectors
criteria were achieved during the two most recent monitoring
periods.  Another commenter (A-90-20:  IV-D-19) recommended
that the EPA should allow an owner or operator to skip to
monitoring connectors every four years if the initial
monitoring shows greater than 0.5 percent leaks.   This
commenter (A-90-20:  IV-D-19)  recommended that the connector
standard use a consistent approach to that in the pump and
valve standard.
     Response;  The EPA agrees that the final rule should
allow owners or operators the flexibility to adopt lower leak
frequency monitoring schedule provided there is documentation
that the criteria for the less frequent monitoring have been
met.  The provisions in §63.174(b) have been revised to
clarify this point.  This clarification does not, however,
allow an owner or operator to elect to use reduced monitoring
frequencies without Method 21 data to document achievement of
lower leak rates for the required periods.  The suggestion by
commenter (A-90-20:  IV-D-19)  that the connector standard
allow an owner or operator to elect to meet requirements of a
later phase would require restructuring the provisions in the
connector standard to specify a lower leak frequency for the
quadrennial monitoring frequency.
     Comment;  One commenter (A-90-19:  IV-D-92)  suggested
that the connector standard incorporate the skipped-period
concepts of §60.483-2 of subpart W.  The commenter  (A-90-19:
IV-D-92) recommended that if the leak rate is less than
2 percent,  the facility could skip one monitoring period, for
semiannual monitoring.  If, after two consecutive monitoring
periods, the leak rate is less than 2 percent, the facility
could skip three monitoring periods, for annual monitoring.
                             5-38

-------
     Response;  Although there is merit to consistency with
concepts in existing programs, it must be noted that the
monitoring frequency and the criteria suggested differ
significantly from the levels agreed to by the negotiating
committee.  For example, the commenter suggested that
quarterly monitoring would be appropriate for units exceeding
the 2 percent leakage rate.  The agreement was for annual
monitoring for leak frequencies greater than 0.5 percent and
biennial monitoring if less than 0.5 percent.  The EPA does
not believe that it would be appropriate to add the
commenter?s suggestion as an optional compliance mechanism.
     Comment:  A number of commenters (A-90-19: IV-D-33;
IV-D-69; IV-D-73; IV-D-77) (A-90-20:  IV-D-19) recommended
that the definition of "inaccessible connectors" be made the
same as the definition of "difficult to monitor" valves.  The
commenters (A-90-19:  IV-D-33; IV-D-69; IV-D-73; IV-D-77)
(A-90-20:  IV-D-19) expressed concern that the requirement to
monitor connectors which can be reached only via a 25-ft
portable scaffold presents safety concerns and is not a cost-
effective means of reducing emissions.  One commenter
(A-90-19: IV-D-77) further noted that piping is generally
constructed above other equipment and the area is covered with
gravel; therefore, rolling scaffolding would not be
appropriate.  The commenter (A-90-19:  IV-D-77) added that,
because of flammability concerns, powered vehicles are not
allowed in these areas.  Therefore, the commenter (A-90-19:
IV-D-77) concluded that the only portable scaffolding that
could be used is field-erected scaffolding.  The commenter
(A-90-19:  IV-D-77) estimated that the cost-effectiveness of
emission reduction achieved if the percent leaking was
0.5 percent would be $2.7 million/ton.
     Similar comments were made by one commenter (A-90-20:
IV-D-10), who was a member of the committee.  This commenter
(A-90-20:  IV-D-10) indicated that the type of portable
scaffolding envisioned was a wheeled scissor lift platform
that would sit on the ground below the monitoring or repair
location.  The commenter (A-90-20:  IV-D-10) reported that the
                             5-39

-------
issue of stable ground was discussed and it was understood
that the scaffolding would not be used on grassed or unstable
stone covered areas below pipelines.  The commenter (A-90-20:
IV-D-10) also noted that raising an individual vertically is
not the only safety issue when trying to reach an inaccessible
connector.  Other safety issues include: (1) danger of
damaging electrical cables and piping; (2)  limitations on
access due to curbs and process equipment spacing; and (3)
dangers of fire and explosions in some process areas.  This
commenter (A-90-20:  IV-D-10) requested clarification of the
term portable scaffold and that issues of safe access be
addressed.
     Response;  The EPA agrees with the commenter (A-90-20:
IV-D-10) that the committee discussions were clear that
implementation of the monitoring provisions was not to
endanger maintenance or monitoring personnel's lives.  The
committee specifically discussed and agreed that use of
scissor lifts on gravel or grass was not intended as well as
use of gas-powered cherry pickers or non-rated electrical
motors in areas with an explosion hazard.
     The EPA believes that these concerns expressed by the
other commenters (A-90-19:  IV-D-33; IV-D-69; IV-D-73;
IV-D-77) (A-90-20:   IV-D-19) are addressed by the provisions
in §63.174(h)(1).  This paragraph has been expanded to include
additional situations where connectors are considered to be
"inaccessible connectors," such as elevating monitoring
personnel two or more meters above a support surface, or
erecting a scaffold.
     The commenters1 (A-90-19:  IV-D-33; IV-D-69; IV-D-73;
IV-D-77) (A-90-20:   IV-D-19) suggestion that the definition  of
"inaccessible" be made the same as "difficult to monitor
valves" is not consistent with the committee's desire to have
connectors monitored where it can be safely conducted.
Revising the definition to be consistent with the "difficult
to monitor valve" definition would not be consistent with the
intent of the negotiated standard since there are situations
                             5-40

-------
where  a wheeled  scissor  lift, platform, or hydraulic
scaffolding  could  be  used  on a paved area within a unit.
     Comment;  One commenter  (A-90-20:  IV-D-19) thought that
the EPA should allow  switching between connector monitoring
alternatives in  §63.174(c)(1)(i) and 63.174(c)(1)(ii) without
the penalty  of more frequent monitoring.
     Response;   The two  connector monitoring options were
provided to  address concerns of some committee members
constituents about the recordkeeping burden of the follow-up
monitoring in §63.174(c)(1)(i).  Since the purpose of the
follow-up monitoring, at least in part, is to establish the
nonrepairable pool, an alternative provision was provided  for
those  companies  willing  to forego the nonrepairable pool in
exchange for less  burdensome administrative costs.
Section 63.174(c)(1)(ii) allows an owner or operator to treat
disturbed connectors  like  any other connector in the unit  for
the purposes of  monitoring in exchange for setting the
nonrepairables pool to zero.  The committee also agreed to
allow  an owner or  operator to switch among alternatives
provided the new alternative is started with an annual
program.  This restriction was included to prevent an owner or
operator selecting the alternative most favorable to him
during that  particular monitoring cycle.  Commenter's
(A-90-20:  IV-D-19) suggestion to remove the penalty of more
frequent monitoring is not appropriate.
     Comment;  One commenter  (A-90-19: IV-D-89) recommended
that a section for difficult-to-monitor connectors should  be
added  to §63.174.   The commenter (A-90-19:  IV-D-89) asserted
that it is not uncommon  to have connectors positioned in such
a manner that elevation  of monitoring personnel more than  2
meters above a support surface would be required.
     Response;   Difficult-to-monitor valves require monitoring
as often as  possible  and at least annually.  The commenter's
concerns are addressed by  the provisions for inaccessible
connectors in §63.174(h) which exempt connectors that are
greater than  2 meters above a support surface and that cannot
be reached using portable  scaffolding.
                             5-41

-------
     Comment;  One conunenter (A-90-20:  IV-D-19)  suggested
that unsafe-to-monitor, unsafe-to-repair and inaccessible
connectors should be excluded from the percent leaking
connectors calculation. The commenter (A-90-20:  IV-D-19)
expressed concerns that because of the infrequent monitoring
performed on these connectors,  unsafe-to-monitor and unsafe-
to-repair connectors results could be very skewed in one
direction or another, giving an unrealistic picture of the
remaining connectors in the process unit.  These connectors
should be excluded from both the number of leaking connectors
(CL) and total connectors (C
-------
available on an as-needed basis or within a fixed time period
after the process units percent leaking valves equals or
exceeds 2 percent.  The commenter  (A-90-19:  IV-D-73) argued
that such an approach is appropriate because the need for this
option may not be apparent within the first year of phase III.
     Response:  The committee restricted the availability of
this QIP due to concerns that it could be used to delay
improving performance.  Some committee members were concerned
that the QIP would never result in improved performance
because the QIP allows the owner or operator to continue
quarterly monitoring and provides 2 years to gather data and
identify better performing equipment.  It was also thought
that the need for the program would be apparent by the first
year of phase III and that sources should have improvements in
performance as experience is gained with the program.  As an
owner or operator may elect to use the QIP during the first
year of phase III regardless of whether the process unit has
2 percent or more leaking valves, every owner or operator of a
source has an opportunity to elect the program.  The
provisions in §63.175(c) also allow an owner or operator to
continue a QIP program after the process unit has fewer than
2 percent leaking valves.
     The EPA would like to recommend that owners or operators
of sources consider developing their own quality
assurance/quality control program that could be used to avoid
election of the formal program in §63.175.  The EPA believes
that a quality control program that is outside the scope of
the provisions in subpart H would have lower recordkeeping
costs and be more flexible to the needs of the facility.
     Comment;  One commenter (A-90-20:  IV-D-19) thought that
the number of valves in the trial evaluation program of a QIP
should be clarified to include only those valves needing
replacing.  Section 63.175(e)(6)(ii) should include only those
valves that have higher leak rates and need to be replaced.
There would be no need to evaluate or replace valves that
work; only those that are inadequate and need replacing.
                             5-43

-------
     Response;  The EPA would like to make clear that the
purpose of the trial evaluation program is to evaluate the
feasibility of using in the process unit subject to the QIP
those valve designs or technologies that others have
identified as having low emission performance.  The
requirement to evaluate the lesser of 1 percent or 20 valves
for single process units (or 1 percent or 50 valves for groups
of process units) is not excessive.  The trial evaluation
program is not directed towards valve replacement, but towards
determining the feasibility of application of other
technologies in the specific process unit.
     Comment;  One commenter (A-90-20:  IV-D-19) argued that
contractors should not be included in the calculation of the
total number of employees at a facility.  The commenter
(A-90-20:  IV-D-19) reasoned that contractor personnel are
temporary, and their number at any given time is variable.
The commenter noted that it is unlikely that a facility would
use contract personnel to get below the cutoff for trial
evaluation—the benefits of this are too small.  The commenter
also argued that the determination should refer to the number
of employees at the facility site, not the entire corporation.
     Response;  The intent of this provision was to reduce the
impact of the requirement on small businesses within this
industry.  Since use of this provision would be easier on
small businesses if it was not necessary to document the
number of temporary contract personnel on site, the
requirement to consider contract personnel was removed.  The
commenter's suggestion that the number of employees be
specified on a facility site basis is not consistent with the
intent to provide some relief for small businesses.  This
suggested edit was not made.
     Comment;  One commenter (A-90-20:  IV-D-19) argued that
the EPA should not require facilities to positively identify
superior performing equipment technologies within 24 months of
the start of the QIP.  This requirement inaccurately assumes
that such a technology can be identified in the first trial
evaluation.  The commenter (A-90-20:  IV-D-19) observed that
                             5-44

-------
this requirement is not consistent with the concept that
performance trials may need to continue for some time.  The
commenter  (A-90-20:  IV-D-19) requested that the EPA remove
this inconsistency.
     Response;  The intent of this requirement is for the
owner or operator to begin trial evaluations of the
technologies that had been identified and not to wait for the
perfect solution.  The provisions of §63.175(e)(6)(iv) require
that the evaluations begin no later than 18 months after the
start of Phase III and be conducted for a minimum of 6 months.
It should be noted that the trial evaluation program is only
required for sites that failed to identify superior performing
technologies during the data analysis phase of the QIP.
5.1.15  §63.176;  Quality Improvement Program for Pumps
     Comment;  One commenter (A-90-20:  IV-D-19) thought that
the number of pumps in the trial evaluation program of a QIP
should be clarified to include only those that need replacing.
Section 63.175(e)(6)(ii) should include only those pumps that
have higher leak rates and need to be replaced.  There would
be no need to evaluate or replace pumps that work; only those
that are inadequate need replacing.
     Response;  The EPA would like to make clear that the
purpose of the trial evaluation program is to evaluate the
feasibility of using in the process unit subject to the QIP
those pump seal designs or technologies that others have
identified as having low emission performance.  The
requirement to evaluate the lesser of 1 percent or 2 pumps for
single process units (or 1 percent or 5 pumps for groups of
process units) is not excessive.  The trial evaluation program
is not directed towards pump replacement, but towards
determining the feasibility of application of other
technologies in the specific process unit.
5.1.16  s63.177;  Alternative Means of Emission Limitation
     Comment;  One commenter (A-90-19:  IV-D-56) urged the EPA
to expand these provisions so that they will apply to all
sections of the rule.   The commenter (A-90-19:  IV-D-56) noted
that by providing alternative compliance options for design,
                             5-45

-------
equipment standards, and work practices, operational
flexibility will be enhanced.
     Response:   The EPA believes that the provisions in
§63.6(g) of the general provisions and the provisions in this
section already provide the flexibility the commenter is
requesting.  Since the commenter did not provide specific
details explaining the unfilled need, it is not possible for
the EPA to address the commenter's concern.
     Comment;  One commenter (A-90-20:  IV-D-19) requested
that the EPA clarify §63.177 to allow use of an alternative
means of emission limitation before the EPA has approved or
disapproved the alternative.  The commenter (A-90-20:
IV-D-19) recommended this change be made to provide
consistency with the Benzene Waste NESHAP  (40 CFR part 61,
subpart FF).
     Response;  The EPA does not believe that it is necessary
to add language to §63.177 that will allow owners or operators
at their own risk to install and operate alternative control
measures, pending approval by the Administrator.  Addition of
such language does not provide owners and operators with any
rights that they did not otherwise have.  As discussed in the
January 7, 1993 FEDERAL REGISTER clarifying amendments to the
Benzene Waste NESHAP (58 FR 3072), if the owner choices to
install or implement an alternative means of emission
limitation prior to approval and it is determined that the
measure does not achieve the emission limitation, the owner
may be cited for noncompliance with the applicable
requirement.
5.1.17  S63.178;  Alternative Means of Emission Limitation for
Batch Processes
     Comment:  Two commenters (A-90-20:  IV-D-20; IV-D-27)
recommended modification of these provisions to allow vacuum
as well as pressure testing.  One commenter (A-90-20:
IV-D-20) stated that in some cases vacuum testing will be
easier because it can be done as part of the inert ing
operation prior to beginning the batch operation.  The other
                              5-46

-------
commenter  (A-90-20:  IV-D-27) reported that vacuum testing can
be accomplished on some systems without adding equipment.
     Response;  The EPA would like to clarify that the
provisions of subpart H do not apply to equipment in vacuum
service.  If a process is operated under a vacuum, there is no
potential for loss of process fluids to the atmosphere through
seal failures in equipment such as valves.  The EPA agrees
that vacuum testing for pressure rise should be allowed.  The
provisions in §63.178(b) and §63.180(f) have been revised to
include vacuum testing.
     Comment:  Two commenters (A-90-20:  IVrD-6; IV-D-27)
stated that a literal interpretation of the proposed
§63.178(b)(1) would require pressure testing each time a seal
is broken during production of the same intermediate or
product and even during a process run.  One commenter
(A-90-20:  IV-D-6) asserted that batch processes in the
pharmaceutical industry, in general, do not have dedicated
batch product trains.  The commenter (A-90-20:  IV-D-6) added
that most process trains are set up to receive feed through a
manifold system which necessitates quick hose connection and
disconnection in order to allow receipt of varying feed
material as the process dictates.  The commenter (A-90-20:
IV-D-6) requested confirmation that the intent of the
provisions in §63.178(b)(1) is to require pressure testing
only when the equipment is reconfigured to produce a different
product or intermediate.  One of the two commenters  (A-90-20:
IV-D-27)  suggested specific language for clarification of the
rule.
     Response:  The intent with these provisions was to
require pressure-testing each time the equipment was
reconfigured for production of another product or
intermediate.  Pressure testing of routine seal breaks, which
are not part of reconfiguration to produce a different
product,  was not envisioned.  If the committee had intended
pressure-testing of routine seal breaks the language in
§63.178(b)(1)  would not have required pressure testing "before
organic HAP is first fed to the equipment".  Additionally, the
                             5-47

-------
last sentence of §63.178(b)(1),  which provides that there
shall be a minimum of one test per year, would not have been
considered necessary if the committee had envisioned routine
seal breaks as being subject to pressure testing.
Section 63.178(b)(l) has been revised to clarify that pressure
testing is not required for routine seal breaks.
     Comment;  One commenter  (A-90-20:  IV-D-19) asserted that
the repair requirements for batch processes failing a pressure
test are not realistic for those process units that contain
material during the test.   The commenter (A-90-20:  IV-D-19)
reported that batch equipment can sometimes be reconfigured
for different products without disturbing all of the equipment
or draining the process lines.  The commenter (A-90-20:
IV-D-19) suggested that the language in §63.178(b)(4) revised
to refer to startup of the process or the second failure.
     Response;  Because the pressure testing requirement is
for new or disturbed equipment the committee language was
drafted assuming that the new equipment would not be in HAP
service at the time of the first test.  The committee thought
that if the reconfigured equipment leaked the problem should
be addressed before the equipment is put into service.  During
the committee discussions it was not envisioned' that this
restriction could be interpreted as applying to other
equipment in the portion of the equipment train that was not
tested.  Since there could be ambiguity regarding the
requirement, the provisions in §63.178(b)(4) have been revised
to clarify the intent.
     Comment;  A commenter (A-90-20:  IV-D-27)  requested
clarification that the proposed pressure testing procedure  is
a minimum requirement, and that equivalent or more rigorous
testing should also be acceptable.  The commenter (A-90-20:
IV-D-27) specifically suggested that testing over a  shorter
test period but at a higher test pressure should also be
permitted.
     Response:  The provisions have been edited to allow
owners or operators the option to conduct the test as
                             5-48

-------
specified at proposal or show that the applicable FDA test has
been conducted.
     Comment;  One commenter  (A-90-19: IV-D-86) supported the
alternative provisions for batch processes in §63.178.  The
commenter (A-90-19:  IV-D-86) endorsed these provisions
because they provide a means of complying without the onerous
recordkeeping requirements of §63.181.
     Response:  While the EPA appreciates the commenters
support, the EPA would like to make clear that there are
recordkeeping requirements associated with §63.178.  These
requirements are located in §63.181(e) of the final subpart H.
Owners or operators of batch processes that comply using the
pressure testing provisions of §63.178(b) are also required to
submit the reports specified in §63.182.
     Comment:  One commenter  (A-90-20:  IV-D-19) suggested
that batch processes that use the pressure testing provisions
of §63.178(b) should be exempt from the requirement for annual
monitoring of the closed vent system.  The commenter  (A-90-20:
IV-D-19) thought that it should be possible to test these
systems during the pressure test of the batch equipment and
monitoring of the closed vent system would be of no benefit.
     Response:  The provisions for closed vent systems were
revised in response to comment and in light of data on the
leak frequency of equipment in these systems,  since the final
provisions of §63.172 do not require an annual monitoring of
closed vent systems, it is not necessary to provide an
exemption for batch equipment that pressure test the system.
See section 5.1.11 of this document for detailed discussion.
     Comment:  One commenter  (A-90-20:  IV-D-19) thought that
the alternative standard for batch provisions should state
that it is permissible for an owner or operator to switch
among the alternatives in paragraphs (b) and (c).
     Response;   As drafted in the proposed standard there is
no restriction on the use of alternatives.  However, due to
the fundamental differences between the two alternatives,
switching among the alternatives does not appear to provide
significant advantages to the source owner or operator.  Since
                             5-49

-------
the committee did not discuss restricting the ability to
switch among the alternative, language has been added to
§63.178(b) to permit that.
5.1.18  S63.179;  Alternative Means of Emission Limitation for
Enclosed-Vented Process Units
     Comment:  Two commenters (A-90-19:  IV-D-26)  (A-90-20:
IV-D-19)  argued that if the equipment is enclosed and vented
as specified, it should be exempt from all monitoring
requirements in the regulation,  as well as the requirements
for visual inspections and equipment standards.
     Response;  The language in §63.179 was drafted to exclude
these units from the monitoring and visual inspection
requirements.  Since these process units must be contained in
a structure operated under a vacuum and vented to a control
device, the EPA agrees that exemption from the equipment
standards would be appropriate.   The language in §63.179 has
been revised to exempt enclosed vented units from the
requirements of §§63.173 through 63.178.  Owners and operators
electing to use this provision are still subject to the
recordkeeping and reporting requirements of §63.181 and
§63.182.
     Comment;  One commenter (A-90-19: IV-D-86) recommended
that enclosed equipment vented to a control device should be
exempted from the monitoring requirements applied to the
equipment.
     Response;  The EPA believes the standard provides this as
an option for compliance for equipment where this option is a
realistic alternative.  Since the commenter did not specify
the type of equipment for which this alternative was desired,
the EPA has no way of determining if changes are necessary or
appropriate.
5.1.19  Repair Procedures
     Comment:  Two commenters (A-90-19:  IV-D-33)  (A-90-20:
TV-D-4) recommended that the proposed  repair intervals.(i.e.,
 irst attempt at repair within 5 calendar days and repair
vithin 15 calendar days) be expressed  in terms of working  days
in order to facilitate scheduling and  reduce overtime
                             5-50

-------
 operating  costs.  These commenters  (A-90-19:  IV-D-33)
 (A-90-20:   IV-D-4) noted that at some facilities maintenance
 staff does not work  24-hour shifts  and some facilities only
 operate on a  5-day work week.  The  commenters (A-90-19:
 IV-D-33)  (A-90-20:   IV-D-4) indicated that their suggestion
 would reduce  the burden of the rule on facilities,
 particularly  small facilities.  One commenter (A-90-19:
 IV-D-33) also expressed the opinion that the definition of
 "days" in  the General Provisions for part 63 would override
 any definitions of days provided in subpart H.  Therefore, the
 commenter  (A-90-19:  IV-D-33) requested that the final rule
 explicitly state each place where the term "working days" can
 be used.
     Response;  The  two commenters  are in effect requesting a
 longer repair interval.  The length of the repair interval
 affects the emission reduction potential of the LDAR program.
 As discussed  in the  preamble to the proposed standard, the
 first attempt at repair is required as soon as practicable and
 no later than 5 days.  Based on experience with the existing
 equipment  leak standards, the EPA believes that 5 days should
 be sufficient time to schedule simple field repairs that do
 not require isolation of the equipment from the process.  The
 standard also provides a 15 day interval for repair of
 equipment  such as valves that do require isolation from the
 process.   Since the  committee did discuss the question of
 repair interval and  retained the 5  calendar day/15 calendar
 day approach  of the  existing standards, it would not be
 appropriate to revise this as suggested.
     The EPA  would like to make clear the relationship between
 the General Provisions to part 63 (subpart A) and subpart H.
 Section 63.1(a)(l) of subpart A provides that individual
 subparts may  include specific definitions in addition to those
 in subpart A  as well as override definitions in subpart A.
 Thus, subpart H could define day to mean working day if that
were appropriate for the provisions in subpart H.  As
discussed above, however, the EPA does not agree that it is
appropriate to use working day in subpart H.
                             5-51

-------
     Comment;  One commenter (A-90-19: IV-D-92) stated that
there would be a serious problem if repair of leaks within
5 days was required.  The commenter (A-90-19:  IV-D-92)
recommended that 15 days be allowed for repair.
     Response:  Subpart H specifies that a first attempt at
repair must be made within 5 days and repair must be completed
within 15 days.  These requirements are consistent with those
in the existing standards for equipment leaks - e.g., SOCMI
equipment leaks NSPS in subpart W of 40 CFR part 60 and
Benzene equipment leaks NESHAP in subpart J of 40 CFR part 61.
     Comment;  One commenter (A-90-20:  IV-D-4) noted that
minor differences among similar standards results in confusion
about the requirements and increases training and other
implementation costs.  The commenter  (A-90-20:  IV-D-4)
suggested that the EPA establish consistent inspection
procedures among the rules.
     Response;  The EPA believes that consistent procedures
are used among the standards.  These procedures are provided
in Method 21 of appendix A to part 60.
5.2  IMPACTS ANALYSIS
     Comments regarding national impacts are in chapter 7.0 of
BID volume 2D.
5.3  APPLICABILITY
5.3.1  Definition of SOCMI
     Comment;  Several commenters (A-90-19:  IV-D-32; IV-D-33;
IV-D-34)  noted that the list of SOCMI processes in subpart H
differs from the list in subpart F.  These commenters
(A-90-19:  IV-D-32; IV-D-33; IV-D-34) requested that the EPA
make the two lists consistent.  One commenter  (A-90-19:
IV-D-33)  noted that the differences between the lists are
principally due to glycol ethers (which are not listed in
subpart H) and the listing of several non-SOCMI products.  The
commenter (A-90-19:  IV-D-33) estimated that there are about
29 SOCMI chemicals that are on one of the two lists  (14
additional chemicals on the SOCMI list in subpart F and 15
chemicals on the list in subpart H).  This commenter  (A-90-19:
IV-D-33)  also stated that there are 7 non-SOCMI chemical
                             5-52

-------
products on the list in subpart F and 13 on the list in
subpart H.  This commenter  (A-90-19:  IV-D-33) recommended
that in order to ease implementation, there should only be one
SOCMI product list, and that list should be provided in
subpart F so that subpart F would specify the applicability of
both subparts G and H for SOCMI processes.
     Response;  The EPA agrees with the commenters and the
final rule has the corrected combined list in section 63.106
of subpart F.  The final SOCMI chemical list consists of 386
products.  See memorandum "Changes to the List of SOCMI
Chemicals in the HON," from Julie Anne Probert, Radian
Corporation, to Janet S. Meyer, EPA/SDB, February 4, 1994 for
a discussion of the specific revisions to the chemical list.
     Comment;  One commenter (A-90-19:  IV-D-33) submitted
corrections to the CAS numbers for five compounds listed in
§63.184 and noted there are duplications of some compounds
through use of synonyms for several compounds.
     Response:  As a result of this comment, the CAS numbers
were corrected and duplicate listings were removed.  These
corrections are discussed in the memorandum "Changes to the
List of SOCMI Chemicals in the HON," from Julie Anne Probert,
Radian Corporation, to Janet S. Meyer, EPA/SDB/ February 4,
1994.
     Comment;  Several commenters (A-90-19: IV-D-34; IV-D-77;
IV-D-97) suggested that various aspects of the applicability
determinations in subparts F and H be made identical.  Several
commenters (A-90-19: IV-D-34; IV-D-77)  (A-90-20:  IV-D-19)
recommended that subpart H be modified to define applicability
on the primary product basis used in subpart F.  These
commenters (A-90-19: IV-D-34; IV-D-77)  (A-90-20:  IV-D-19)
thought the language in subpart F is much more definitive than
the language in subpart H.  Another commenter  (A-90-19:
IV-D-97) recommended that subpart H refer to subpart F
applicability and the specific applicability section in
subpart H be deleted.
     Response;  As discussed in BID volume 2D, the
applicability for SOCMI is now defined in subpart F.  All
                             5-53

-------
specification of applicability has been removed from
subpart H.
5.3.2  Definition of Non-SOCMI Processes
     Comment;   Several commenters (A-90-19: IV-D-33; IV-D-71;
IV-D-77; IV-D-86; IV-D-97) argued the seven non-SOCMI
processes should be regulated separately from the SOCMI
processes.  The commenters (A-90-19:  IV-D-33; IV-D-71;
IV-D-77; IV-D-86; IV-D-97) argued that covering the non-SOCMI
categories in more than one section 112(d)  standard would
result in potentially overlapping regulations, which would
increase the burden to the regulated industry and create
confusion for regulatory agencies.  One commenter (A-90-19:
IV-D-77) suggested that, at the very least, the EPA should
move the non-SOCMI applicability requirements to the subpart
reserved for these source categories.
     Another commenter  (A-90-19:  IV-D-34)  argued that
regulation of the non-SOCMI categories should be deferred
until rules for the other emission points in these categories
are issued.  The commenter (A-90-19:  IV-D-34) thought that
this would avoid piecemeal regulation of the non-SOCMI
categories.
     Response;  The EPA agrees with the commenters1
suggestions for reorganization of the regulatory provisions
into separate subparts.  The applicability provisions for the
non-SOCMI processes have been placed in subpart I.  Subpart I
refers to subpart H for the substantive requirements.
Subpart I also contains revised definitions for some of the
non-SOCMI processes.  The definitions were revised to
specifically identify the intended processes and to directly
correspond to the categories listed in the source category
list (57 FR 31576).  The EPA expects that these changes will
minimize the possibility for confusion and development of
overlapping regulations.  The EPA does not agree with the
suggestion to defer the applicability of the equipment leak
standard to the non-SOCMI processes until requirements for the
other emission points are issued.  This change suggested by
                             5-54

-------
the commenters is not consistent with the spirit of the
negotiated agreement.
     Comment;  One commenter  (A-90-19: IV-D-77) recommended
that the final rule should include appropriate criteria for
determining applicability for the non-SOCMI categories.  The
commenter (A-90-19:  IV-D-77) asserted that the criteria used
for SOCMI processes are not well suited to specialty,
intermediate, or pharmaceutical processes where the intended
product can be a relatively minor portion of the total mass
output.  In other cases, the commenter (A-90-19:  IV-D-77)
agreed that the SOCMI criteria were appropriate.
     Response:  The EPA agrees with the commenter that in the
case of several of the non-SOCMI categories the criteria for
the SOCMI processes are not appropriate.  The applicability
for the non-SOCMI processes is now specified in subpart I.
Thus, there should no longer be any implication that the
criteria used for SOCMI apply to these processes.
     Comment:  One commenter  (A-90-19: IV-D-92) requested
clarification of the definitions of "miscellaneous butadiene
use" and "chlorinated hydrocarbon use."  The commenter
(A-90-19:  IV-D-92) stressed that the definitions implied
production.
     Response;  The EPA agrees with the commenter.  Because
the processes grouped under these terms used these chemicals
to produce other chemicals, the definitions have been revised
to more specifically identify these processes.
     Commenti  One commenter  (A-90-19:  IV-D-26) argued that
the proposed definition of "pesticide production" defines a
source category that is broader than any category in the
agricultural chemical production industry grouping identified
in the List of Source Categories (57 FR 31576), or in the
Schedule for Promulgation of Emission Standards  (57 FR 44147).
The commenter (A-90-19:  IV-D-26) noted that only five of the
source categories listed in 57 FR 31576 or 57 FR 44147 were
identified as being specifically regulated by the HON
equipment leak standard.  The commenter (A-90-19:  IV-D-26)
reasoned that the proposed definition would include processes
                              5-55

-------
that are not producing one of the listed agricultural chemical
products identified as a source category under §112(c) of the
Act.  The commenter (A-90-19:  IV-D-26)  noted that if the EPA
wants to update the source category list, then the EPA must
comply with statutory requirements to include only categories
of major sources and area sources where a finding of adverse
health effects has been made.  The commenter (A-90-19:
IV-D-26) recommended that the definition of pesticide
production in subpart H be modified to regulate equipment
leaks only from the production of Captafol, Captan,
Chlorothalonil, Dacthal™, and Tordon™.  Another comment
(A-90-19:  IV-D-lll) questioned whether formulation of
pesticide products was subject to the negotiated rule.
     Response;  The definition for pesticide production in
proposed subpart H was developed in the regulatory negotiation
before creation of the source category list.  The source
category list and schedule (57 FR 31576 and 57 FR 44147)
identifies the pesticide production processes that the EPA had
information on at the time of the negotiations on the
equipment leak standard.  Therefore, subpart H is being
applied only to the five production processes identified in
the source category list (production of Captafol, Captan,
Chlorothalonil, Dacthal™, and Tordon™).  As the EPA obtains
information on other pesticide processes, these processes will
be added to the source category list in the future and
standards will be developed for these categories.  During
these separate rulemakings the EPA will consider what the
appropriate control requirements should be for the new source
categories.  It should not be assumed that the provisions of
subpart H will be automatically applied to these categories.
     The EPA would like to emphasize that the pesticide
processes subject to the negotiated rule are producers of the
active pesticide ingredient, and not formulators.
     Comment;  One commenter (A-90-19:  IV-D-lll) requested
that a definition of "processing aid" be added to subpart H.
                              5-56

-------
     Response;  The term  "processing aid" is not used in the
 final  subparts  F, H, or I.  Therefore a definition was not
 necessary.
     Comment;   One commenter  (A-90-19: IV-D-71) suggested that
 if  it  is the EPA's intent that subpart H regulate fugitive
 emissions from  all organic operations, then the list of
 processes must  be expanded and the EPA must provide an
 additional opportunity for public comment.
     Response;  It is not the EPA's intent that subpart H
 regulate all organic operations.  Subpart H is intended only
 to  regulate listed items.  The inclusion of the seven
 non-SOCMI processes was a product of the regulatory
 negotiation process.
 5.3.3  Equipment Subnect  to Subpart H
     Comment;   One commenter  (A-90-19: IV-D-92) argued that
 subpart H should only apply to equipment in greater than
 10  percent VHAP service.  The commenter  (A-90-19:  IV-D-92)
 indicated that  such a change would limit the program to
 volatile organic materials, thereby eliminating inorganic and
 low-volatility  substances such as polynuclear aromatic
 compounds.
     Response:  The commenter provided no supporting arguments
 regarding why the applicability level should be increased from
 5 percent to 10 percent.  The applicability level of 5 percent
 total VHAP was  selected as part of the negotiation process.
 Because the list of chemicals being regulated in the rule does
 not include inorganic compounds, the rule does not apply to
 them.
     cofflpep^;  one commenter  (A-90-20: IV-D-27) supported the
 EPA's distinction between process and utility lines (57 FR
 62660)  and recommended that subpart H include language that
 explicitly exempts utilities and non-process lines.  The
 commenter supported this  distinction because heat transfer
 lines may include ethylene glycol and other VHAP's but
 typically will be impossible to monitor because of insulation.
     Response;  The EPA believes that explicit exemption of
nonprocess lines and utilities should not be necessary.
                              5-57

-------
First, applicability of the standard is clearly centered
around process operations.  Second, even if there were
confusion regarding the first point, there would be no effect
because insulated equipment is exempt from the monitoring
requirement.
5.3.4  Miscellaneous
     Comment;  One commenter (A-90-19:  IV-D-26) recommended
that the language in §63.100(b)(2) be amended to clarify that
the rule only applies to major sources.  The commenter
(A-90-19:  IV-D-26) noted that the EPA had not made a finding
of adverse health effects for any of the categories identified
in the December 31, 1992 proposal.  Thus, without the
clarification that this rule applies only to major sources,
the proposed rule far exceeds the statutory authority provided
to the EPA under the Act.
     Response;  The standard applies to major sources only
which is clearly stated in the final rule §63.100(b)(3) of
subpart F.  As indicated in the preamble to the proposed rule,
the EPA did not have information on area sources in SOCMI and
lacked the basis for making a finding of adverse health
effects.
     Comment:  One commenter (A-90-19: IV-D-77)' requested that
the final rule provide some consideration of operating
schedule for flexible operation units.  The commenter
(A-90-19:  IV-D-77) noted that in some cases, the specified
monitoring frequency is greater than the frequency of product
changes within the chemical manufacturing process.
     Response;  The EPA believes that the commenter's concern
has been addressed through two sets of changes made in the
final rule.  One of the two changes is to the applicability
provisions for flexible operation units.  The other change is
a clarification of the monitoring requirements in subpart H.
     First, in developing the final rule, the EPA reevaluated
the proposed approach that a flexible operation unit would be
subject only during those times the unit was producing a
chemical listed in subpart F.  Based on public comments and
possible interaction with provisions of 112(g), the EPA
                             5-58

-------
decided that the proposed approach would complicate compliance
demonstrations for sources and enforcement agencies.  The
final rule requires flexible operations be assigned to a
specific chemical manufacturing process based on anticipated
use of the unit.  For existing sources, the assignment is to
be based on the expected use over the first five years
following promulgation of the standard.  For new sources, the
assignment is to be based on the expected use in the first
five years after initial startup.
     Second, §63.180 of subpart H now specifies that
monitoring can be performed under several different
conditions.  These conditions are when the equipment is:
(1) in organic HAP service; (2) in use with an acceptable
surrogate VOC which is not an organic HAP; or (3) in use with
any other detectable gas or vapor.
     Comment:  One commenter (A-90-19:  IV-K-80), who operates
a pharmaceutical process that would be subject to subpart H,
requested that the EPA allow compliance extensions for process
changes to eliminate use of HAP's.  The commenter (A-90-19:
IV-K-80) plans to startup the new process about 3 months after
they would have to comply with the monitoring requirements  in
subpart H.
     Response;  During the negotiations, there was no
discussion of possibility of process changes to eliminate use
of HAP's and how that should be considered in terms of
compliance schedule.  The regulations provide up to 3 years to
install controls for other types of emission points, e.g.
process vents.  Three years were not given to equipment leak
sources because there is no large capital expenditure;
implementation of the rule consists primarily of labor costs.
However, an analogous situation to installation of controls
with regard to equipment leaks would be a process change that
would eliminate the HAP's of concern that would cause a source
to be subject to the HON.  Such changes can typically involve
significant reworking of equipment, e.g., reconfiguration,
replacement, addition, or removal of equipment.  Process
changes take more time than a few months to effect, especially
                             5-59

-------
considering planning, approval of permits, and in some cases
approval by the Food and Drug Administration or other
government entities.
     In at least one situation, the Agency is aware of a
pharmaceutical source making such a process change that would
be subject to the HON for only a few months, until the change
has been completed.  The cost associated with monitoring and
recordkeeping for the short amount of time the HON would apply
has been estimated by the company to be about $100,000.  This
estimate is probably not unreasonable, but even if it were
less, the Agency believes requiring a company to meet a
standard for about 3 months (the length of time the company
estimated that it would be subject to the HON prior to
eliminating the HAP of concern),  and incur the associated
costs, is not warranted.  It would not be fair to penalize a
company that will be eliminating HAP emissions subject to the
HON, if allowing a little more time to make the change is
necessary.
     In order to provide comparable time for a source to
complete a process change, the Agency will allow that source
up to 18 months after promulgation.  This is a reasonable
amount of time, as opposed to 3 years, given that sources that
would have been subject to the equipment leak rules had an
earlier opportunity to view the draft rules (56 FR 9315;
March 6, 1991).   Eighteen months is also the same schedule
that the last group of process units become subject to the
rules, in effect moving those units undergoing a process
change to the last group.
     This extension would be conditional upon the source
successfully making the process change prior to (18 months
after promulgation).  This would be necessary so that a source
could not claim it was making a change, and then not do so,
only to gain an advantage over competitors by delaying the
cost associated with the rule.  A source would have no more
than 18 months after promulgation of the HON to complete the
process change.  If, after 18 months, the source is still
subject to the HON, it would be retroactively subject to
                              5-60

-------
penalties  for the time during the extension that it did not
comply with the HON.
The source would still have to file an initial report with the
EPA, describing the process change, the HAP's eliminated, and
the expected date of cessation of operation of the current
process.
5.4  COMPLIANCE DEMONSTRATIONS
5.4.1  Monitoring
     Comment;  Several commenters (A-90-19:  IV-D-14; IV-D-15;
IV-D-17; IV-D-18; IV-D-19; IV-D-20; IV-D-23; IV-D-24; IV-D-25;
IV-D-27; IV-D-28; IV-D-40; IV-D-42) suggested that
§63.180(b)(4) be modified to allow the use of the calibration
gas best suited to the detector technology in use.  These
commenters (A-90-19:  IV-D-14; IV-D-15; IV-D-17; IV-D-18;
IV-D-19; IV-D-20; IV-D-23; IV-D-24; IV-D-25; IV-D-27; IV-D-28)
contended that the requirement to use methane as the
calibration gas precluded the use of photoionization detectors
for measuring total volatile organics.  Furthermore, the
commenters (A-90-19:  IV-D-14; IV-D-15; IV-D-17; IV-D-18;
IV-D-19; IV-D-20; IV-D-23; IV-D-24; IV-D-25; IV-D-27; IV-D-28)
indicated that the specification of methane was contradictory
because the rule requires adjustment for response factors
greater than 3.
     Response;  Method 21 has not been revised.  Methane is
specified as the calibration gas to identify methane as the
sole reference gas for determining the response factors.  If
the leak detection instrument to be used has a poor response
or no response to methane, then section 3.1.2 of Method 21 can
still be cited as a valid justification for choosing a
different calibration gas.  If a calibration gas other than
methane is used, then instrument readings must be converted to
a methane basis, as specified in section 3.2 of Method 21.
This conversion factor would be the ratio of the molecular
weight of the calibration gas to the molecular weight of
methane.
     Comment:   One commenter (A-90-20:  IV-D-19) argued that
the requirement in §63.180(b)(4)(i) for 0.2 ppm zero air will
                             5-61

-------
increase the cost of compliance without any benefit.  The
commenter (A-90-20:  IV-D-19) recommended that the
specifications of Method 21 for zero air be used instead.
     Response:   The final standard uses the same definition
of zero air as Method 21, 10 ppm.  This requirement is better
suited to the intent of the requirement for leak detection.
The proposed requirement of 0.2 ppm reflected an earlier
committee concern with obtaining better data for bagging
studies.
     Comment:  Two commenters (A-90-19: IV-D-77) (A-90-20:
IV-D-19) requested that the final rule allow background
subtraction for leak determinations, where leaks are defined
as 500 ppm by Method 21.  Failure to allow subtraction of
background makes the standard more stringent than existing
standards.
     Response;  The final standard has been revised to allow
subtraction of background readings.  This adjustment was
overlooked at the time the negotiated rule was drafted and
does not represent an attempt to make the standard more
stringent.
     Comment;  Several commenters  (A-90-19: IV-D-73; IV-D-92)
(A-90-20:  IV-D-19) recommen^-i deleting • ie requirement to
calculate and use response f:   rs to dete.raine whether
components are leaking.  The _ .Jimenters (A-90-19:  IV-D-73;
IV-D-92) (A-90-20:  IV-D-19) stated that most response factors
are less than 3, incremental benefit is small, and the program
complication is immense.  Several commenters  (A-90-19:
IV-D-69; IV-D-73) added that the equation in the Federal
Register notice was incorrect and differed from the version
agreed to by the committee.  One of these commenters  (A-90-19:
IV-D-69) noted that the committee equation is only applicable
to gas mixtures.  For mixtures of liquids, the equation must
be adjusted by the vapor pressure of each Component.
     Response;  The EPA reevaluated this p  posed requirement
in light of these comments and the experiei je gained by
chemical plants that have implemented the negotiated rule.
Experience with the correction for response factors has shown
                             5-62

-------
the proposed provisions to be significantly more burdensome
than originally anticipated.  Specifically, several facilities
have reported that instead of quantifying 4 to 5 compounds in
some streams they are quantifying 50 to 100 compounds.  In
some cases, owners or operators have elected to correct all
instrument readings by the highest response factor for any
compound  in the process rather than undertake the effort
associated with the stream specific corrections.
     The  EPA also reviewed the reasons the Committee
originally considered requiring adjustment of screening values
by response factors.  The response factor adjustment
originated in Committee discussions on studies to improve the
emission  estimates.  (Response factors are used to correct
instrument readings to indicate actual concentrations for
developing emission estimates.)  The appropriateness of
adjusting screening values in the leak detection and repair
provisions was not considered.  These adjustments will not
change the emission reductions achieved from implementing the
standard.  Therefore, the EPA believes that eliminating this
provision does not change the effect of the standard and
preserves the Committee's intent of minimizing unproductive
effort.
     The  final standard requires the owner or operator to use
a monitoring instrument that meets the specifications of
Method 21 of appendix A of 40 CFR part 60.  The proposed
provisions in section 63.180(b)(6) have been removed from the
standard, although response factor adjustments are still
allowed in cases where no acceptable instrument is available.
It has been clarified that the leak definitions are expressed
in terms of total VOC, and not speciated concentration
readings.
     Comment:  One commenter (A-90-20:  IV-D-19) observed that
Method 21 requires fresh daily calibration gas unless a longer
shelf life can be demonstrated.  The commenter (A-90-20:
IV-D-19), thus, thought that this requirement makes use of
many process streams as calibration gases impractical as the
stream would have to be resampled and analyzed daily.
                             5-63

-------
     Response;  The requirement for response factor adjustment
has been removed from the final standard.  As noted above, the
instrument must still meet the specifications of Method 21.
5.4.2  Test Methods
     Comment ;   One commenter (A-90-20: IV-D-27)  suggested
modifying §63. 180(b) (4) (iii) to allow calibration at more than
one point.  The commenter noted the proposed requirement fails
to take advantage of the flexibility of many monitoring
instruments.  The commenter, thus, thought the proposed
requirement would result in delays and additional labor for
monitoring equipment such as agitators.
     Response :  The provisions do not prohibit calibrating at
more than one point.  The provisions require that one of the
calibration points be at the leak definition.
     Comment :   One commenter (A-90-20:  IV-D-19) observed that
Method 21 requires determination of no detectable emissions
using 2.5 percent of the specified leak definition.  The
commenter noted that for a 500 ppm leak definition, no
detectable emissions equates to 25 ppm, which is beyond the
accuracy of most monitoring instruments.
     Response:  The EPA is uncertain as to the nature of the
commenter 's concern since subpart H does not use the concept
of no detectable emissions.  Thus, while the commenter may be
correct that 25 ppm may not be accurately measured by some or
many instruments, the concern is not relevant to the
requirements of subpart H.
5.5  RECORDKEEPING AND REPORTING
5.5.1  General
               One commenter (A-90-20:  IV-D-4) expressed
concern that differences among reporting and recordkeeping
requirements makes implementation of rules more difficult.
This commenter (A-90-20:  IV-D-4) also expressed concern that
some States require entirely different reporting procedures;
thus, two-records are required and implementation is more
difficult.
     Response :  The EPA does not believe that there should be
a problem with differences among rules reporting requirements.
                             5-64

-------
 First,  subpart  H  provides that equipment subject to existing
 NSPS  or NESHAP  and  subpart H need only comply with the
 provisions  of subpart H.   Thus, there is only one reporting
 and recordkeeping system required.  Second, subpart H provides
 that  if the state requirements provides the same information,
 those reports can be submitted for the reports required by
 subpart H.   Because subpart H does not specify a required
 format  for  records  or reports, the EPA thinks that two sets of
 records will not  be necessary.
      Comment;   Several commenters (A-90-19:  IV-D-33; IV-D-69)
 (A-90-20:   IV-D-20) thought the proposed recordkeeping
 requirements were extremely burdensome.  One commenter
 (A-90-20:   IV-D-20) recommended that the recordkeeping section
 be modified to  eliminate all the detailed cross-referencing
 that  LDAR requires, unless a QIP becomes required at the
 facility.   Another  commenter (A-90-19:  IV-D-69) recommended
 that  the initial  report should only contain a statement that
 the facility is subject to subpart H.  This commenter
 (A-90-19:   IV-D-69) suggested that all other information be
 made  available  upon request in a reasonable time.  The
 benefits of this  approach were that the permit authority would
 receive complete  information at the time they need it for
 compliance  determination and the facility would not have to
 prepare a report  at a time when they are starting to implement
 subpart H.   This  commenter (A-90-19:  IV-D-69) also suggested
 that  facilities in  compliance with subpart H should only
 submit  a statement  that the monitoring has occurred and they
 are in  compliance.
      Response:  The EPA agrees that, to the extent possible,
 reporting requirements for subpart H should be simplified and
 the required records clarified.  However, the EPA does not
 think that  some of  the commenter's suggestions can be adopted
 because the  records are essential to verify compliance or
 required by  the operating permit rule.  For example, there
 needs to be  some  record of units and equipment subject to the
 standard.  Therefore, to streamline reporting requirements and
minimize potential  confusion, the following changes have been
                             5-65

-------
made to the proposed requirements:  (1) submit the compliance
notification report no later than 90 days after the
applicability date for the group of chemical processes and the
periodic reports every 6 months thereafter; (2) allow the
source to adjust the reporting schedule to combine the
periodic reports for subpart H with those for subpart G once
the source comes into compliance with the provisions in
subpart G; and (3) consolidation of several special reports
into the periodic report.
     The final standard does not allow annual reporting as
requested by several commenters.  Although many components may
be monitored on an annual basis, the monitoring frequency for
pumps and agitators is monthly and annual reporting would not
be consistent with the reporting system in subpart G.  The
operating permit provisions of the Act, section 504(a), also
requires reporting to be at least semiannual.  Additionally,
unless all units at a source are on the same monitoring
schedule, the source would have to submit multiple annual
reports.  The EPA believes that consolidating the reporting
into semiannual reports for subparts G and H is more efficient
for both industry and enforcement agencies.
     In light of these comments, the EPA reevaluated the
proposed requirements to ensure that only those records and
reports essential for enforcement of the standard are
required.  This review showed that some of the commenters1
concerns arose from a lack of clarity in the proposed standard
regarding actual records required for some of the provisions
and other concerns arose from overlapping or duplicative
requirements.  It was also determined that some provisions
should be redrafted to be compatible with computerized data
management systems and the revised provisions would still
provide the information necessary to demonstrate compliance.
Examples of such changes include: allowing a source to
maintain on file a written procedure outlining the conditions
for delay of repair and requiring certain records only for
nonautomated systems.  The EPA also reviewed the proposed
                             5-66

-------
 standard to  identify  implied recordkeeping requirements and to
 specify all  the required records  in §63.181.
     Comment;  One commenter (A-90-19: IV-D-104) urged the EPA
 to  include provisions for reduced recordkeeping and monitoring
 requirements after a  facility demonstrates sustained
 compliance over a reasonable period of time.  The commenter
 (A-90-19:  IV-D-104)  was very concerned with the cumulative
 impact of all the Clean Air Act requirements.
     Response;  In regards to monitoring, the proposed and
 final equipment leak  provisions provide for a reduced
 monitoring frequency  for good performance.  With a reduced
 monitoring frequency, less records are required.  The EPA has
 taken reasonable steps to minimize and remove unnecessary or
 redundant records in  the final rule.  The monitoring and
 recordkeeping requirements that are in the final rule are what
 the EPA considers necessary for ensuring compliance with the
 rule.
     Comment;  One commenter (A-90-20: IV-D-27) urged that
 subpart H explicitly  authorize use of computer readable
 records.  The commenter noted that subpart G includes computer
 readable records in the definition of continuous records.
     Response:  In the proposed subpart F it specified that
 records could be computer records.  In the final subparts F
 and I, this  provision remains with additional clarifying
 language stating:
     Records may be maintained in hard copy or computer-
     readable form including, but not limited to, on
     paper, microfilm, computer,  floppy disk, magnetic
     tape,  or microfiche.
 5.5.2  Batch Processes
     comment;  One commenter (A-90-19: IV-D-77) recommended
that batch processes  which operate infrequently should be
subject to the requirement for the initial report but should
be subject to only annual reporting thereafter.  The commenter
 (A-90-19:   IV-D-77)  asserted that the proposed semiannual
reporting frequency would impose  an undue burden.
     Response;   The title V provisions for operating permits
requires semi-annual  reports, therefore it would not reduce
                             5-67

-------
the burden to the industry for the equipment leaks provisions
to require annual reports.  The EPA has minimized the records
as much as possible so that the information contained in the
semi-annual reports is less burdensome.
     Comment:  One commenter (A-90-19: IV-D-77) recommended
that the final rule should explicitly state that the
requirements in §63.181(f)(5) for records of any visual,
auditory, or olfactory evidence of fluid loss are to be made
during the pressure test.
     Response;  This has been clarified in the final rule.
     Comment;  One commenter (A-90-20:  IV-D-27) urged the EPA
to reconsider the recordkeeping requirements for batch
processes to reduce the burden.  A specific concern noted by
the commenter was the difficulties of assigning equipment to a
particular process unit because of frequent reconfigurations.
The commenter thought that these changes would make it
difficult to manage the database.
     Response;  The requirements for batch processes have been
revised in the final rule.  The requirements now direct that
records be maintained for product or product code and
information on whether testing has been conducted when process
has been reconfigured for producing a different product.
Owners or operators no longer have to have records of the
individual items of equipment.
5.6  WORDING OF PROVISIONS
     Comment;  One commenter (A-90-19: IV-D-74) requested that
the definition of "equipment leak" in subpart F should be
deleted and that in each instance where the term is used, the
rule should be rephrased to clarify the intended meaning.  The
commenter (A-90-19:  IV-D-74) suggested this edit because the
definition in subpart F refers to the equipment subject to
subpart H and does not recognize that the equipment is only
leaking if monitored above a specified concentration.
     Response;   Subpart F has been clarified  in the final
rule to specify the equipment that is affected and not refer
to "equipment leaks."
                             5-68

-------
     Comment;  One conunenter (A-90-19: IV-D-92) recommended
that "equipment in heavy liquid service" should be defined as
"equipment not in VHAP gas/vapor service or VHAP light liquid
service."
     Response;  The definition of "in heavy liquid service" in
the final rule is as follows:
     In heavy liquid service means that a piece of
     equipment in organic hazardous air pollutant service
     is not in gas/vapor service or in light liquid
     service.
This was the same definition as at proposal, except that at
proposal the term "in VHAP service" was used instead of "in
HAP service."
     Comment;  One commenter (A-90-19: IV-D-86) indicated that
the EPA should clarify §63.160(c) to clearly state that only
the parenthetically listed HAP's are subject to the
requirements.
     Response;  The EPA feels that the applicability of the
subpart H requirements is already clear.  Thus, no changes
have been made to §63.160(c).
     Comment;  One commenter (A-90-19: IV-D-95) contended that
the tables in §63.178, §63.183, and §63.184 should be
identified by a number and a title.
     Response;  All tables in the final rule are treated
consistently and according to the requirements set forth by
the Office of the Federal Register.
                             5-69

-------
                                     TECHNICAL REPORT DATA
                             (Please read Instructions on the reverse be/ore completing/
i REPORT NO
  EPA-453/R-94-003a
                                                              3 RECIPIENT'S ACCESSION NO
4. TITLE AND SUBTITLE
  Hazardous Air Pollutant laiaaiona from Procaa* Onita in th«'synthetic Organic
  Chanical Manufacturing Industry—Background Information for Final Standard*
  Volume 2Ai
            5. REPORT DATE
              March 1994
             int« on Procaaa Vanti, Storage Vaaaala, Tranifar Operation*,
          and Equipment Laaka
            6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
                                                              8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
  Office of  Air Quality  Planning and Standards
  U.S. Environmental Protection Agency
  Research Triangle Park,  North Carolina  27711
                                                              10. PROGRAM ELEMENT NO.
             11. CONTRACT/GRANT NO.

               68D10117
12. SPONSORING AGENCY NAME AND ADDRESS
  Director,  Office of  Air Quality  Planning and Standards
  Office of  Air and Radiation
  U.S. Environmental Protection Agency
  Research Triangle Park, North Carolina  27711
                                                              13. TYPE OF REPORT AND PERIOD COVERED
             14. SPONSORING AGENCY CODE

               EPA/200/04
15. SUPPLEMENTARY NOTES
16. ABSTRACT

  A  final rule  for the regulation of emissions of  organic hazardous air
  pollutants  (HAP's) from  chemical processes of the synthetic  organic chemical
  manufacturing industry  (SOCMI) is being promulgated under  the authority of
  sections 112,  114, 116,  and 301 of the Clean Air Act, as amended in 1990.  The
  emission standards were  proposed in  the Federal  Register on  December 31, 1992
   (57  FR 62608).   Public hearings were held.  A supplemental notice was
  published in  the Federal Register on October 15,  1993 (58  FR 53478).   This
  volume of the background information document summarizes all comments  and
  presents the  agency's responses on process vents, storage  vessels, transfer
  operations, and equipment leaks.
17.
                                 KEY WORDS AND DOCUMENT ANALYSIS
                   DESCRIPTORS
                                                b.lDENTIFIERS/OPEN ENDED TERMS
                           c. COSATI Field/Group
  Air pollution
  Pollution control
  SOCMI
  Hazardous air pollutant
  National impacts
Air pollution control
18. DISTRIBUTION STATEMENT
                                                 19. SECURITY CLASS /Tins Report/
                           21  NO. OF PAGES
                                 225
                                                20. SECURITY CLASS i Tins page!
                                                   UNCLASSIFIED
                                                                             22 PRICE
EPA Form"2220-l (R«v. 4-77)   PREVIOUS EDITION is OBSOLETE

-------